| { |
| "best_metric": 2.9464175701141357, |
| "best_model_checkpoint": "bert_base_lda_book/checkpoint-590000", |
| "epoch": 25.0, |
| "eval_steps": 10000, |
| "global_step": 593075, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.021076592336551025, |
| "grad_norm": 2.3989920616149902, |
| "learning_rate": 5e-06, |
| "loss": 11.2086, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.04215318467310205, |
| "grad_norm": 1.2999556064605713, |
| "learning_rate": 1e-05, |
| "loss": 8.9924, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.06322977700965308, |
| "grad_norm": 1.2827718257904053, |
| "learning_rate": 1.5e-05, |
| "loss": 8.3436, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.0843063693462041, |
| "grad_norm": 1.1802339553833008, |
| "learning_rate": 2e-05, |
| "loss": 8.2208, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.10538296168275513, |
| "grad_norm": 1.1368786096572876, |
| "learning_rate": 2.5e-05, |
| "loss": 8.1601, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.12645955401930617, |
| "grad_norm": 0.8442460298538208, |
| "learning_rate": 3e-05, |
| "loss": 8.12, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.14753614635585718, |
| "grad_norm": 0.9719900488853455, |
| "learning_rate": 3.5e-05, |
| "loss": 8.0902, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.1686127386924082, |
| "grad_norm": 1.1849809885025024, |
| "learning_rate": 4e-05, |
| "loss": 8.063, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.18968933102895924, |
| "grad_norm": 1.080064058303833, |
| "learning_rate": 4.5e-05, |
| "loss": 8.0388, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.21076592336551025, |
| "grad_norm": 0.9970459342002869, |
| "learning_rate": 5e-05, |
| "loss": 8.023, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.2318425157020613, |
| "grad_norm": 0.856968104839325, |
| "learning_rate": 5.500000000000001e-05, |
| "loss": 8.0118, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.25291910803861234, |
| "grad_norm": 1.046067476272583, |
| "learning_rate": 6e-05, |
| "loss": 7.9937, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.27399570037516335, |
| "grad_norm": 2.123368740081787, |
| "learning_rate": 6.500000000000001e-05, |
| "loss": 7.9866, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.29507229271171437, |
| "grad_norm": 1.1102254390716553, |
| "learning_rate": 7e-05, |
| "loss": 7.9532, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.3161488850482654, |
| "grad_norm": 1.1079896688461304, |
| "learning_rate": 7.500000000000001e-05, |
| "loss": 7.9334, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.3372254773848164, |
| "grad_norm": 1.0515861511230469, |
| "learning_rate": 8e-05, |
| "loss": 7.901, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.35830206972136747, |
| "grad_norm": 1.810900330543518, |
| "learning_rate": 8.5e-05, |
| "loss": 7.8012, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.3793786620579185, |
| "grad_norm": 1.421406626701355, |
| "learning_rate": 9e-05, |
| "loss": 7.7031, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.4004552543944695, |
| "grad_norm": 1.0557781457901, |
| "learning_rate": 9.5e-05, |
| "loss": 7.6749, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.4215318467310205, |
| "grad_norm": 0.9890698194503784, |
| "learning_rate": 0.0001, |
| "loss": 7.6611, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.4215318467310205, |
| "eval_accuracy": 0.16256870790620534, |
| "eval_loss": 7.483117580413818, |
| "eval_runtime": 178.1094, |
| "eval_samples_per_second": 677.707, |
| "eval_steps_per_second": 7.063, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.4426084390675716, |
| "grad_norm": 1.2449626922607422, |
| "learning_rate": 9.99142477382841e-05, |
| "loss": 7.6465, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.4636850314041226, |
| "grad_norm": 1.1824065446853638, |
| "learning_rate": 9.98284954765682e-05, |
| "loss": 7.6388, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.4847616237406736, |
| "grad_norm": 1.2166095972061157, |
| "learning_rate": 9.97427432148523e-05, |
| "loss": 7.635, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.5058382160772247, |
| "grad_norm": 1.1044520139694214, |
| "learning_rate": 9.96569909531364e-05, |
| "loss": 7.623, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.5269148084137757, |
| "grad_norm": 1.2120144367218018, |
| "learning_rate": 9.957123869142048e-05, |
| "loss": 7.6164, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.5479914007503267, |
| "grad_norm": 1.1285841464996338, |
| "learning_rate": 9.94854864297046e-05, |
| "loss": 7.6104, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.5690679930868777, |
| "grad_norm": 1.0910948514938354, |
| "learning_rate": 9.939973416798869e-05, |
| "loss": 7.6016, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.5901445854234287, |
| "grad_norm": 0.9991886615753174, |
| "learning_rate": 9.931398190627278e-05, |
| "loss": 7.5969, |
| "step": 14000 |
| }, |
| { |
| "epoch": 0.6112211777599798, |
| "grad_norm": 0.9672591090202332, |
| "learning_rate": 9.922822964455688e-05, |
| "loss": 7.5885, |
| "step": 14500 |
| }, |
| { |
| "epoch": 0.6322977700965308, |
| "grad_norm": 1.1728878021240234, |
| "learning_rate": 9.914247738284099e-05, |
| "loss": 7.5855, |
| "step": 15000 |
| }, |
| { |
| "epoch": 0.6533743624330818, |
| "grad_norm": 0.9827332496643066, |
| "learning_rate": 9.905672512112507e-05, |
| "loss": 7.5753, |
| "step": 15500 |
| }, |
| { |
| "epoch": 0.6744509547696328, |
| "grad_norm": 1.147050142288208, |
| "learning_rate": 9.897097285940917e-05, |
| "loss": 7.5748, |
| "step": 16000 |
| }, |
| { |
| "epoch": 0.6955275471061839, |
| "grad_norm": 1.045847773551941, |
| "learning_rate": 9.888522059769327e-05, |
| "loss": 7.5665, |
| "step": 16500 |
| }, |
| { |
| "epoch": 0.7166041394427349, |
| "grad_norm": 1.331439733505249, |
| "learning_rate": 9.879946833597737e-05, |
| "loss": 7.5599, |
| "step": 17000 |
| }, |
| { |
| "epoch": 0.737680731779286, |
| "grad_norm": 0.8554311394691467, |
| "learning_rate": 9.871371607426147e-05, |
| "loss": 7.5527, |
| "step": 17500 |
| }, |
| { |
| "epoch": 0.758757324115837, |
| "grad_norm": 0.8436219692230225, |
| "learning_rate": 9.862796381254556e-05, |
| "loss": 7.5588, |
| "step": 18000 |
| }, |
| { |
| "epoch": 0.779833916452388, |
| "grad_norm": 0.9435406923294067, |
| "learning_rate": 9.854221155082965e-05, |
| "loss": 7.5494, |
| "step": 18500 |
| }, |
| { |
| "epoch": 0.800910508788939, |
| "grad_norm": 0.8975987434387207, |
| "learning_rate": 9.845645928911376e-05, |
| "loss": 7.5502, |
| "step": 19000 |
| }, |
| { |
| "epoch": 0.82198710112549, |
| "grad_norm": 0.9397053122520447, |
| "learning_rate": 9.837070702739786e-05, |
| "loss": 7.5472, |
| "step": 19500 |
| }, |
| { |
| "epoch": 0.843063693462041, |
| "grad_norm": 0.7552306056022644, |
| "learning_rate": 9.828495476568194e-05, |
| "loss": 7.54, |
| "step": 20000 |
| }, |
| { |
| "epoch": 0.843063693462041, |
| "eval_accuracy": 0.1640321339439997, |
| "eval_loss": 7.389256954193115, |
| "eval_runtime": 181.3799, |
| "eval_samples_per_second": 665.487, |
| "eval_steps_per_second": 6.936, |
| "step": 20000 |
| }, |
| { |
| "epoch": 0.864140285798592, |
| "grad_norm": 1.2272006273269653, |
| "learning_rate": 9.819920250396604e-05, |
| "loss": 7.5327, |
| "step": 20500 |
| }, |
| { |
| "epoch": 0.8852168781351432, |
| "grad_norm": 0.746410608291626, |
| "learning_rate": 9.811345024225015e-05, |
| "loss": 7.5288, |
| "step": 21000 |
| }, |
| { |
| "epoch": 0.9062934704716942, |
| "grad_norm": 0.6792486906051636, |
| "learning_rate": 9.802769798053424e-05, |
| "loss": 7.529, |
| "step": 21500 |
| }, |
| { |
| "epoch": 0.9273700628082452, |
| "grad_norm": 0.8895508050918579, |
| "learning_rate": 9.794194571881834e-05, |
| "loss": 7.5202, |
| "step": 22000 |
| }, |
| { |
| "epoch": 0.9484466551447962, |
| "grad_norm": 0.9248941540718079, |
| "learning_rate": 9.785619345710244e-05, |
| "loss": 7.5179, |
| "step": 22500 |
| }, |
| { |
| "epoch": 0.9695232474813472, |
| "grad_norm": 0.8147923946380615, |
| "learning_rate": 9.777044119538653e-05, |
| "loss": 7.5144, |
| "step": 23000 |
| }, |
| { |
| "epoch": 0.9905998398178982, |
| "grad_norm": 0.8717142343521118, |
| "learning_rate": 9.768468893367063e-05, |
| "loss": 7.5105, |
| "step": 23500 |
| }, |
| { |
| "epoch": 1.0116764321544494, |
| "grad_norm": 0.8675404191017151, |
| "learning_rate": 9.759893667195473e-05, |
| "loss": 7.5078, |
| "step": 24000 |
| }, |
| { |
| "epoch": 1.0327530244910004, |
| "grad_norm": 0.886225700378418, |
| "learning_rate": 9.751318441023883e-05, |
| "loss": 7.5054, |
| "step": 24500 |
| }, |
| { |
| "epoch": 1.0538296168275514, |
| "grad_norm": 0.7221285104751587, |
| "learning_rate": 9.742743214852293e-05, |
| "loss": 7.5043, |
| "step": 25000 |
| }, |
| { |
| "epoch": 1.0749062091641024, |
| "grad_norm": 0.8174595832824707, |
| "learning_rate": 9.734167988680703e-05, |
| "loss": 7.4942, |
| "step": 25500 |
| }, |
| { |
| "epoch": 1.0959828015006534, |
| "grad_norm": 0.9026414752006531, |
| "learning_rate": 9.725592762509111e-05, |
| "loss": 7.4915, |
| "step": 26000 |
| }, |
| { |
| "epoch": 1.1170593938372044, |
| "grad_norm": 0.7532177567481995, |
| "learning_rate": 9.717017536337522e-05, |
| "loss": 7.4885, |
| "step": 26500 |
| }, |
| { |
| "epoch": 1.1381359861737554, |
| "grad_norm": 1.0045020580291748, |
| "learning_rate": 9.708442310165932e-05, |
| "loss": 7.4934, |
| "step": 27000 |
| }, |
| { |
| "epoch": 1.1592125785103065, |
| "grad_norm": 0.9671968221664429, |
| "learning_rate": 9.69986708399434e-05, |
| "loss": 7.4869, |
| "step": 27500 |
| }, |
| { |
| "epoch": 1.1802891708468575, |
| "grad_norm": 0.8180074095726013, |
| "learning_rate": 9.69129185782275e-05, |
| "loss": 7.4903, |
| "step": 28000 |
| }, |
| { |
| "epoch": 1.2013657631834085, |
| "grad_norm": 0.6831914186477661, |
| "learning_rate": 9.68271663165116e-05, |
| "loss": 7.4809, |
| "step": 28500 |
| }, |
| { |
| "epoch": 1.2224423555199595, |
| "grad_norm": 0.880370020866394, |
| "learning_rate": 9.67414140547957e-05, |
| "loss": 7.4782, |
| "step": 29000 |
| }, |
| { |
| "epoch": 1.2435189478565105, |
| "grad_norm": 0.8444699048995972, |
| "learning_rate": 9.66556617930798e-05, |
| "loss": 7.4724, |
| "step": 29500 |
| }, |
| { |
| "epoch": 1.2645955401930615, |
| "grad_norm": 0.8642208576202393, |
| "learning_rate": 9.656990953136388e-05, |
| "loss": 7.477, |
| "step": 30000 |
| }, |
| { |
| "epoch": 1.2645955401930615, |
| "eval_accuracy": 0.16630146985552097, |
| "eval_loss": 7.3398661613464355, |
| "eval_runtime": 180.7043, |
| "eval_samples_per_second": 667.975, |
| "eval_steps_per_second": 6.962, |
| "step": 30000 |
| }, |
| { |
| "epoch": 1.2856721325296125, |
| "grad_norm": 0.815250813961029, |
| "learning_rate": 9.6484157269648e-05, |
| "loss": 7.4688, |
| "step": 30500 |
| }, |
| { |
| "epoch": 1.3067487248661636, |
| "grad_norm": 0.7102167010307312, |
| "learning_rate": 9.639840500793209e-05, |
| "loss": 7.472, |
| "step": 31000 |
| }, |
| { |
| "epoch": 1.3278253172027146, |
| "grad_norm": 0.7698110342025757, |
| "learning_rate": 9.631265274621618e-05, |
| "loss": 7.4681, |
| "step": 31500 |
| }, |
| { |
| "epoch": 1.3489019095392658, |
| "grad_norm": 0.705014705657959, |
| "learning_rate": 9.622690048450028e-05, |
| "loss": 7.4698, |
| "step": 32000 |
| }, |
| { |
| "epoch": 1.3699785018758166, |
| "grad_norm": 1.0288499593734741, |
| "learning_rate": 9.614114822278439e-05, |
| "loss": 7.4651, |
| "step": 32500 |
| }, |
| { |
| "epoch": 1.3910550942123678, |
| "grad_norm": 0.8654530644416809, |
| "learning_rate": 9.605539596106847e-05, |
| "loss": 7.4673, |
| "step": 33000 |
| }, |
| { |
| "epoch": 1.4121316865489189, |
| "grad_norm": 0.752121090888977, |
| "learning_rate": 9.596964369935257e-05, |
| "loss": 7.4628, |
| "step": 33500 |
| }, |
| { |
| "epoch": 1.4332082788854699, |
| "grad_norm": 0.8441411256790161, |
| "learning_rate": 9.588389143763667e-05, |
| "loss": 7.4586, |
| "step": 34000 |
| }, |
| { |
| "epoch": 1.4542848712220209, |
| "grad_norm": 0.8407287001609802, |
| "learning_rate": 9.579813917592077e-05, |
| "loss": 7.459, |
| "step": 34500 |
| }, |
| { |
| "epoch": 1.475361463558572, |
| "grad_norm": 0.6961820125579834, |
| "learning_rate": 9.571238691420487e-05, |
| "loss": 7.4578, |
| "step": 35000 |
| }, |
| { |
| "epoch": 1.496438055895123, |
| "grad_norm": 0.7345203757286072, |
| "learning_rate": 9.562663465248896e-05, |
| "loss": 7.4532, |
| "step": 35500 |
| }, |
| { |
| "epoch": 1.517514648231674, |
| "grad_norm": 0.8560628890991211, |
| "learning_rate": 9.554088239077306e-05, |
| "loss": 7.4479, |
| "step": 36000 |
| }, |
| { |
| "epoch": 1.538591240568225, |
| "grad_norm": 0.7769248485565186, |
| "learning_rate": 9.545513012905716e-05, |
| "loss": 7.4507, |
| "step": 36500 |
| }, |
| { |
| "epoch": 1.559667832904776, |
| "grad_norm": 0.8969734907150269, |
| "learning_rate": 9.536937786734126e-05, |
| "loss": 7.4536, |
| "step": 37000 |
| }, |
| { |
| "epoch": 1.580744425241327, |
| "grad_norm": 0.7307207584381104, |
| "learning_rate": 9.528362560562534e-05, |
| "loss": 7.4489, |
| "step": 37500 |
| }, |
| { |
| "epoch": 1.601821017577878, |
| "grad_norm": 0.8283652067184448, |
| "learning_rate": 9.519787334390946e-05, |
| "loss": 7.4475, |
| "step": 38000 |
| }, |
| { |
| "epoch": 1.622897609914429, |
| "grad_norm": 1.1097286939620972, |
| "learning_rate": 9.511212108219355e-05, |
| "loss": 7.4422, |
| "step": 38500 |
| }, |
| { |
| "epoch": 1.64397420225098, |
| "grad_norm": 0.7026250958442688, |
| "learning_rate": 9.502636882047764e-05, |
| "loss": 7.4427, |
| "step": 39000 |
| }, |
| { |
| "epoch": 1.6650507945875312, |
| "grad_norm": 0.8573418855667114, |
| "learning_rate": 9.494061655876174e-05, |
| "loss": 7.4437, |
| "step": 39500 |
| }, |
| { |
| "epoch": 1.686127386924082, |
| "grad_norm": 0.7835220694541931, |
| "learning_rate": 9.485486429704585e-05, |
| "loss": 7.4478, |
| "step": 40000 |
| }, |
| { |
| "epoch": 1.686127386924082, |
| "eval_accuracy": 0.16667498174961573, |
| "eval_loss": 7.311859130859375, |
| "eval_runtime": 181.2687, |
| "eval_samples_per_second": 665.896, |
| "eval_steps_per_second": 6.94, |
| "step": 40000 |
| }, |
| { |
| "epoch": 1.7072039792606333, |
| "grad_norm": 0.7612914443016052, |
| "learning_rate": 9.476911203532993e-05, |
| "loss": 7.4356, |
| "step": 40500 |
| }, |
| { |
| "epoch": 1.728280571597184, |
| "grad_norm": 0.7828889489173889, |
| "learning_rate": 9.468335977361403e-05, |
| "loss": 7.4381, |
| "step": 41000 |
| }, |
| { |
| "epoch": 1.7493571639337353, |
| "grad_norm": 0.84151691198349, |
| "learning_rate": 9.459760751189813e-05, |
| "loss": 7.4393, |
| "step": 41500 |
| }, |
| { |
| "epoch": 1.770433756270286, |
| "grad_norm": 0.8556711077690125, |
| "learning_rate": 9.451185525018223e-05, |
| "loss": 7.4393, |
| "step": 42000 |
| }, |
| { |
| "epoch": 1.7915103486068373, |
| "grad_norm": 0.70616215467453, |
| "learning_rate": 9.442610298846633e-05, |
| "loss": 7.434, |
| "step": 42500 |
| }, |
| { |
| "epoch": 1.8125869409433881, |
| "grad_norm": 0.8368786573410034, |
| "learning_rate": 9.434035072675043e-05, |
| "loss": 7.4303, |
| "step": 43000 |
| }, |
| { |
| "epoch": 1.8336635332799394, |
| "grad_norm": 0.7400949001312256, |
| "learning_rate": 9.425459846503451e-05, |
| "loss": 7.4288, |
| "step": 43500 |
| }, |
| { |
| "epoch": 1.8547401256164904, |
| "grad_norm": 0.8047006726264954, |
| "learning_rate": 9.416884620331862e-05, |
| "loss": 7.4278, |
| "step": 44000 |
| }, |
| { |
| "epoch": 1.8758167179530414, |
| "grad_norm": 1.044785737991333, |
| "learning_rate": 9.408309394160272e-05, |
| "loss": 7.0078, |
| "step": 44500 |
| }, |
| { |
| "epoch": 1.8968933102895924, |
| "grad_norm": 1.1563804149627686, |
| "learning_rate": 9.39973416798868e-05, |
| "loss": 6.5114, |
| "step": 45000 |
| }, |
| { |
| "epoch": 1.9179699026261434, |
| "grad_norm": 1.287293791770935, |
| "learning_rate": 9.39115894181709e-05, |
| "loss": 6.3161, |
| "step": 45500 |
| }, |
| { |
| "epoch": 1.9390464949626944, |
| "grad_norm": 1.1442444324493408, |
| "learning_rate": 9.382583715645502e-05, |
| "loss": 6.1582, |
| "step": 46000 |
| }, |
| { |
| "epoch": 1.9601230872992454, |
| "grad_norm": 0.9973781704902649, |
| "learning_rate": 9.37400848947391e-05, |
| "loss": 5.8725, |
| "step": 46500 |
| }, |
| { |
| "epoch": 1.9811996796357965, |
| "grad_norm": 1.0666022300720215, |
| "learning_rate": 9.36543326330232e-05, |
| "loss": 5.7306, |
| "step": 47000 |
| }, |
| { |
| "epoch": 2.0022762719723475, |
| "grad_norm": 1.1702784299850464, |
| "learning_rate": 9.35685803713073e-05, |
| "loss": 5.6357, |
| "step": 47500 |
| }, |
| { |
| "epoch": 2.0233528643088987, |
| "grad_norm": 1.1498465538024902, |
| "learning_rate": 9.34828281095914e-05, |
| "loss": 5.5454, |
| "step": 48000 |
| }, |
| { |
| "epoch": 2.0444294566454495, |
| "grad_norm": 1.110739827156067, |
| "learning_rate": 9.339707584787549e-05, |
| "loss": 5.4689, |
| "step": 48500 |
| }, |
| { |
| "epoch": 2.0655060489820007, |
| "grad_norm": 1.2224382162094116, |
| "learning_rate": 9.331132358615959e-05, |
| "loss": 5.3556, |
| "step": 49000 |
| }, |
| { |
| "epoch": 2.0865826413185515, |
| "grad_norm": 1.216245412826538, |
| "learning_rate": 9.322557132444369e-05, |
| "loss": 5.242, |
| "step": 49500 |
| }, |
| { |
| "epoch": 2.1076592336551028, |
| "grad_norm": 1.284480094909668, |
| "learning_rate": 9.313981906272779e-05, |
| "loss": 5.084, |
| "step": 50000 |
| }, |
| { |
| "epoch": 2.1076592336551028, |
| "eval_accuracy": 0.498186168509195, |
| "eval_loss": 4.603658199310303, |
| "eval_runtime": 201.5943, |
| "eval_samples_per_second": 598.757, |
| "eval_steps_per_second": 6.24, |
| "step": 50000 |
| }, |
| { |
| "epoch": 2.1287358259916536, |
| "grad_norm": 1.3405102491378784, |
| "learning_rate": 9.305406680101189e-05, |
| "loss": 4.922, |
| "step": 50500 |
| }, |
| { |
| "epoch": 2.149812418328205, |
| "grad_norm": 1.0608346462249756, |
| "learning_rate": 9.296831453929597e-05, |
| "loss": 4.8002, |
| "step": 51000 |
| }, |
| { |
| "epoch": 2.1708890106647556, |
| "grad_norm": 1.1604127883911133, |
| "learning_rate": 9.288256227758008e-05, |
| "loss": 4.7125, |
| "step": 51500 |
| }, |
| { |
| "epoch": 2.191965603001307, |
| "grad_norm": 1.0957660675048828, |
| "learning_rate": 9.279681001586418e-05, |
| "loss": 4.6499, |
| "step": 52000 |
| }, |
| { |
| "epoch": 2.2130421953378576, |
| "grad_norm": 1.1827034950256348, |
| "learning_rate": 9.271105775414827e-05, |
| "loss": 4.588, |
| "step": 52500 |
| }, |
| { |
| "epoch": 2.234118787674409, |
| "grad_norm": 1.1235727071762085, |
| "learning_rate": 9.262530549243236e-05, |
| "loss": 4.5419, |
| "step": 53000 |
| }, |
| { |
| "epoch": 2.2551953800109596, |
| "grad_norm": 1.1212998628616333, |
| "learning_rate": 9.253955323071646e-05, |
| "loss": 4.4952, |
| "step": 53500 |
| }, |
| { |
| "epoch": 2.276271972347511, |
| "grad_norm": 1.2187979221343994, |
| "learning_rate": 9.245380096900056e-05, |
| "loss": 4.4579, |
| "step": 54000 |
| }, |
| { |
| "epoch": 2.2973485646840617, |
| "grad_norm": 1.2245925664901733, |
| "learning_rate": 9.236804870728466e-05, |
| "loss": 4.4111, |
| "step": 54500 |
| }, |
| { |
| "epoch": 2.318425157020613, |
| "grad_norm": 1.1893351078033447, |
| "learning_rate": 9.228229644556876e-05, |
| "loss": 4.3809, |
| "step": 55000 |
| }, |
| { |
| "epoch": 2.339501749357164, |
| "grad_norm": 1.2642035484313965, |
| "learning_rate": 9.219654418385286e-05, |
| "loss": 4.344, |
| "step": 55500 |
| }, |
| { |
| "epoch": 2.360578341693715, |
| "grad_norm": 1.288195013999939, |
| "learning_rate": 9.211079192213695e-05, |
| "loss": 4.3252, |
| "step": 56000 |
| }, |
| { |
| "epoch": 2.381654934030266, |
| "grad_norm": 1.3157896995544434, |
| "learning_rate": 9.202503966042105e-05, |
| "loss": 4.2936, |
| "step": 56500 |
| }, |
| { |
| "epoch": 2.402731526366817, |
| "grad_norm": 1.1476653814315796, |
| "learning_rate": 9.193928739870514e-05, |
| "loss": 4.2643, |
| "step": 57000 |
| }, |
| { |
| "epoch": 2.423808118703368, |
| "grad_norm": 1.2981932163238525, |
| "learning_rate": 9.185353513698925e-05, |
| "loss": 4.2408, |
| "step": 57500 |
| }, |
| { |
| "epoch": 2.444884711039919, |
| "grad_norm": 1.1152125597000122, |
| "learning_rate": 9.176778287527335e-05, |
| "loss": 4.2186, |
| "step": 58000 |
| }, |
| { |
| "epoch": 2.4659613033764702, |
| "grad_norm": 1.0963436365127563, |
| "learning_rate": 9.168203061355743e-05, |
| "loss": 4.2002, |
| "step": 58500 |
| }, |
| { |
| "epoch": 2.487037895713021, |
| "grad_norm": 1.042327880859375, |
| "learning_rate": 9.159627835184153e-05, |
| "loss": 4.1798, |
| "step": 59000 |
| }, |
| { |
| "epoch": 2.5081144880495723, |
| "grad_norm": 1.3668285608291626, |
| "learning_rate": 9.151052609012564e-05, |
| "loss": 4.1649, |
| "step": 59500 |
| }, |
| { |
| "epoch": 2.529191080386123, |
| "grad_norm": 1.184714674949646, |
| "learning_rate": 9.142477382840973e-05, |
| "loss": 4.1458, |
| "step": 60000 |
| }, |
| { |
| "epoch": 2.529191080386123, |
| "eval_accuracy": 0.6153193638711232, |
| "eval_loss": 3.8417911529541016, |
| "eval_runtime": 209.1626, |
| "eval_samples_per_second": 577.092, |
| "eval_steps_per_second": 6.014, |
| "step": 60000 |
| }, |
| { |
| "epoch": 2.5502676727226743, |
| "grad_norm": 1.1496775150299072, |
| "learning_rate": 9.133902156669383e-05, |
| "loss": 4.1253, |
| "step": 60500 |
| }, |
| { |
| "epoch": 2.571344265059225, |
| "grad_norm": 1.169020414352417, |
| "learning_rate": 9.125326930497792e-05, |
| "loss": 4.1138, |
| "step": 61000 |
| }, |
| { |
| "epoch": 2.5924208573957763, |
| "grad_norm": 1.1891062259674072, |
| "learning_rate": 9.116751704326202e-05, |
| "loss": 4.0965, |
| "step": 61500 |
| }, |
| { |
| "epoch": 2.613497449732327, |
| "grad_norm": 1.3287880420684814, |
| "learning_rate": 9.108176478154612e-05, |
| "loss": 4.0815, |
| "step": 62000 |
| }, |
| { |
| "epoch": 2.6345740420688784, |
| "grad_norm": 1.1200419664382935, |
| "learning_rate": 9.09960125198302e-05, |
| "loss": 4.0771, |
| "step": 62500 |
| }, |
| { |
| "epoch": 2.655650634405429, |
| "grad_norm": 1.1130471229553223, |
| "learning_rate": 9.091026025811432e-05, |
| "loss": 4.0583, |
| "step": 63000 |
| }, |
| { |
| "epoch": 2.6767272267419804, |
| "grad_norm": 1.1523027420043945, |
| "learning_rate": 9.082450799639842e-05, |
| "loss": 4.048, |
| "step": 63500 |
| }, |
| { |
| "epoch": 2.6978038190785316, |
| "grad_norm": 1.0300532579421997, |
| "learning_rate": 9.07387557346825e-05, |
| "loss": 4.0373, |
| "step": 64000 |
| }, |
| { |
| "epoch": 2.7188804114150824, |
| "grad_norm": 1.1078014373779297, |
| "learning_rate": 9.06530034729666e-05, |
| "loss": 4.0246, |
| "step": 64500 |
| }, |
| { |
| "epoch": 2.739957003751633, |
| "grad_norm": 1.0599591732025146, |
| "learning_rate": 9.05672512112507e-05, |
| "loss": 4.0123, |
| "step": 65000 |
| }, |
| { |
| "epoch": 2.7610335960881844, |
| "grad_norm": 1.1839988231658936, |
| "learning_rate": 9.04814989495348e-05, |
| "loss": 4.0036, |
| "step": 65500 |
| }, |
| { |
| "epoch": 2.7821101884247357, |
| "grad_norm": 1.1682463884353638, |
| "learning_rate": 9.03957466878189e-05, |
| "loss": 3.9969, |
| "step": 66000 |
| }, |
| { |
| "epoch": 2.8031867807612865, |
| "grad_norm": 1.102096676826477, |
| "learning_rate": 9.030999442610299e-05, |
| "loss": 3.9848, |
| "step": 66500 |
| }, |
| { |
| "epoch": 2.8242633730978377, |
| "grad_norm": 1.1180987358093262, |
| "learning_rate": 9.022424216438709e-05, |
| "loss": 3.9704, |
| "step": 67000 |
| }, |
| { |
| "epoch": 2.8453399654343885, |
| "grad_norm": 1.1765016317367554, |
| "learning_rate": 9.013848990267119e-05, |
| "loss": 3.9625, |
| "step": 67500 |
| }, |
| { |
| "epoch": 2.8664165577709397, |
| "grad_norm": 1.1318899393081665, |
| "learning_rate": 9.005273764095529e-05, |
| "loss": 3.9539, |
| "step": 68000 |
| }, |
| { |
| "epoch": 2.8874931501074905, |
| "grad_norm": 1.0751798152923584, |
| "learning_rate": 8.996698537923937e-05, |
| "loss": 3.9527, |
| "step": 68500 |
| }, |
| { |
| "epoch": 2.9085697424440418, |
| "grad_norm": 1.091174840927124, |
| "learning_rate": 8.988123311752348e-05, |
| "loss": 3.9362, |
| "step": 69000 |
| }, |
| { |
| "epoch": 2.9296463347805926, |
| "grad_norm": 1.0489026308059692, |
| "learning_rate": 8.979548085580758e-05, |
| "loss": 3.9333, |
| "step": 69500 |
| }, |
| { |
| "epoch": 2.950722927117144, |
| "grad_norm": 1.064015507698059, |
| "learning_rate": 8.970972859409167e-05, |
| "loss": 3.9219, |
| "step": 70000 |
| }, |
| { |
| "epoch": 2.950722927117144, |
| "eval_accuracy": 0.6467732328373411, |
| "eval_loss": 3.6448755264282227, |
| "eval_runtime": 210.2336, |
| "eval_samples_per_second": 574.152, |
| "eval_steps_per_second": 5.984, |
| "step": 70000 |
| }, |
| { |
| "epoch": 2.9717995194536946, |
| "grad_norm": 1.0569514036178589, |
| "learning_rate": 8.962397633237576e-05, |
| "loss": 3.9166, |
| "step": 70500 |
| }, |
| { |
| "epoch": 2.992876111790246, |
| "grad_norm": 1.1297013759613037, |
| "learning_rate": 8.953822407065988e-05, |
| "loss": 3.9079, |
| "step": 71000 |
| }, |
| { |
| "epoch": 3.0139527041267966, |
| "grad_norm": 1.1361900568008423, |
| "learning_rate": 8.945247180894396e-05, |
| "loss": 3.9016, |
| "step": 71500 |
| }, |
| { |
| "epoch": 3.035029296463348, |
| "grad_norm": 1.1925541162490845, |
| "learning_rate": 8.936671954722806e-05, |
| "loss": 3.893, |
| "step": 72000 |
| }, |
| { |
| "epoch": 3.0561058887998986, |
| "grad_norm": 1.107208013534546, |
| "learning_rate": 8.928096728551216e-05, |
| "loss": 3.891, |
| "step": 72500 |
| }, |
| { |
| "epoch": 3.07718248113645, |
| "grad_norm": 1.1451010704040527, |
| "learning_rate": 8.919521502379626e-05, |
| "loss": 3.8785, |
| "step": 73000 |
| }, |
| { |
| "epoch": 3.0982590734730007, |
| "grad_norm": 1.137524127960205, |
| "learning_rate": 8.910946276208035e-05, |
| "loss": 3.8761, |
| "step": 73500 |
| }, |
| { |
| "epoch": 3.119335665809552, |
| "grad_norm": 1.0663111209869385, |
| "learning_rate": 8.902371050036445e-05, |
| "loss": 3.8639, |
| "step": 74000 |
| }, |
| { |
| "epoch": 3.140412258146103, |
| "grad_norm": 1.0170351266860962, |
| "learning_rate": 8.893795823864855e-05, |
| "loss": 3.8651, |
| "step": 74500 |
| }, |
| { |
| "epoch": 3.161488850482654, |
| "grad_norm": 1.1180704832077026, |
| "learning_rate": 8.885220597693265e-05, |
| "loss": 3.8579, |
| "step": 75000 |
| }, |
| { |
| "epoch": 3.182565442819205, |
| "grad_norm": 1.1297398805618286, |
| "learning_rate": 8.876645371521675e-05, |
| "loss": 3.8488, |
| "step": 75500 |
| }, |
| { |
| "epoch": 3.203642035155756, |
| "grad_norm": 1.5382742881774902, |
| "learning_rate": 8.868070145350083e-05, |
| "loss": 3.846, |
| "step": 76000 |
| }, |
| { |
| "epoch": 3.224718627492307, |
| "grad_norm": 1.1835236549377441, |
| "learning_rate": 8.859494919178493e-05, |
| "loss": 3.8422, |
| "step": 76500 |
| }, |
| { |
| "epoch": 3.245795219828858, |
| "grad_norm": 1.0558855533599854, |
| "learning_rate": 8.850919693006904e-05, |
| "loss": 3.8338, |
| "step": 77000 |
| }, |
| { |
| "epoch": 3.2668718121654092, |
| "grad_norm": 1.0952107906341553, |
| "learning_rate": 8.842344466835313e-05, |
| "loss": 3.8266, |
| "step": 77500 |
| }, |
| { |
| "epoch": 3.28794840450196, |
| "grad_norm": 1.1452620029449463, |
| "learning_rate": 8.833769240663723e-05, |
| "loss": 3.8212, |
| "step": 78000 |
| }, |
| { |
| "epoch": 3.3090249968385113, |
| "grad_norm": 1.100569725036621, |
| "learning_rate": 8.825194014492132e-05, |
| "loss": 3.8191, |
| "step": 78500 |
| }, |
| { |
| "epoch": 3.330101589175062, |
| "grad_norm": 1.13761305809021, |
| "learning_rate": 8.816618788320542e-05, |
| "loss": 3.8151, |
| "step": 79000 |
| }, |
| { |
| "epoch": 3.3511781815116133, |
| "grad_norm": 1.0493966341018677, |
| "learning_rate": 8.808043562148952e-05, |
| "loss": 3.8097, |
| "step": 79500 |
| }, |
| { |
| "epoch": 3.372254773848164, |
| "grad_norm": 1.051387906074524, |
| "learning_rate": 8.799468335977362e-05, |
| "loss": 3.8057, |
| "step": 80000 |
| }, |
| { |
| "epoch": 3.372254773848164, |
| "eval_accuracy": 0.6641025593965367, |
| "eval_loss": 3.533684015274048, |
| "eval_runtime": 210.847, |
| "eval_samples_per_second": 572.482, |
| "eval_steps_per_second": 5.966, |
| "step": 80000 |
| }, |
| { |
| "epoch": 3.3933313661847153, |
| "grad_norm": 1.116188645362854, |
| "learning_rate": 8.790893109805772e-05, |
| "loss": 3.802, |
| "step": 80500 |
| }, |
| { |
| "epoch": 3.414407958521266, |
| "grad_norm": 1.0311115980148315, |
| "learning_rate": 8.782317883634182e-05, |
| "loss": 3.7957, |
| "step": 81000 |
| }, |
| { |
| "epoch": 3.4354845508578173, |
| "grad_norm": 1.0306453704833984, |
| "learning_rate": 8.773742657462591e-05, |
| "loss": 3.7901, |
| "step": 81500 |
| }, |
| { |
| "epoch": 3.456561143194368, |
| "grad_norm": 1.1646533012390137, |
| "learning_rate": 8.765167431291e-05, |
| "loss": 3.7859, |
| "step": 82000 |
| }, |
| { |
| "epoch": 3.4776377355309194, |
| "grad_norm": 1.1253823041915894, |
| "learning_rate": 8.756592205119411e-05, |
| "loss": 3.7811, |
| "step": 82500 |
| }, |
| { |
| "epoch": 3.4987143278674706, |
| "grad_norm": 1.126400351524353, |
| "learning_rate": 8.748016978947821e-05, |
| "loss": 3.7791, |
| "step": 83000 |
| }, |
| { |
| "epoch": 3.5197909202040214, |
| "grad_norm": 1.074724793434143, |
| "learning_rate": 8.73944175277623e-05, |
| "loss": 3.7764, |
| "step": 83500 |
| }, |
| { |
| "epoch": 3.540867512540572, |
| "grad_norm": 1.145216703414917, |
| "learning_rate": 8.730866526604639e-05, |
| "loss": 3.7737, |
| "step": 84000 |
| }, |
| { |
| "epoch": 3.5619441048771234, |
| "grad_norm": 1.1301367282867432, |
| "learning_rate": 8.72229130043305e-05, |
| "loss": 3.7689, |
| "step": 84500 |
| }, |
| { |
| "epoch": 3.5830206972136747, |
| "grad_norm": 1.0151203870773315, |
| "learning_rate": 8.713716074261459e-05, |
| "loss": 3.7636, |
| "step": 85000 |
| }, |
| { |
| "epoch": 3.6040972895502255, |
| "grad_norm": 1.0575908422470093, |
| "learning_rate": 8.705140848089869e-05, |
| "loss": 3.7623, |
| "step": 85500 |
| }, |
| { |
| "epoch": 3.6251738818867767, |
| "grad_norm": 1.057502269744873, |
| "learning_rate": 8.696565621918278e-05, |
| "loss": 3.7524, |
| "step": 86000 |
| }, |
| { |
| "epoch": 3.6462504742233275, |
| "grad_norm": 1.0766448974609375, |
| "learning_rate": 8.687990395746688e-05, |
| "loss": 3.7525, |
| "step": 86500 |
| }, |
| { |
| "epoch": 3.6673270665598787, |
| "grad_norm": 1.0963397026062012, |
| "learning_rate": 8.679415169575098e-05, |
| "loss": 3.7487, |
| "step": 87000 |
| }, |
| { |
| "epoch": 3.6884036588964295, |
| "grad_norm": 1.0679845809936523, |
| "learning_rate": 8.670839943403508e-05, |
| "loss": 3.7431, |
| "step": 87500 |
| }, |
| { |
| "epoch": 3.7094802512329808, |
| "grad_norm": 1.0979522466659546, |
| "learning_rate": 8.662264717231916e-05, |
| "loss": 3.7415, |
| "step": 88000 |
| }, |
| { |
| "epoch": 3.7305568435695315, |
| "grad_norm": 1.118834376335144, |
| "learning_rate": 8.653689491060328e-05, |
| "loss": 3.7379, |
| "step": 88500 |
| }, |
| { |
| "epoch": 3.751633435906083, |
| "grad_norm": 1.0471783876419067, |
| "learning_rate": 8.645114264888737e-05, |
| "loss": 3.7347, |
| "step": 89000 |
| }, |
| { |
| "epoch": 3.7727100282426336, |
| "grad_norm": 1.0941870212554932, |
| "learning_rate": 8.636539038717146e-05, |
| "loss": 3.7269, |
| "step": 89500 |
| }, |
| { |
| "epoch": 3.793786620579185, |
| "grad_norm": 1.026440143585205, |
| "learning_rate": 8.627963812545556e-05, |
| "loss": 3.7243, |
| "step": 90000 |
| }, |
| { |
| "epoch": 3.793786620579185, |
| "eval_accuracy": 0.6761338118918749, |
| "eval_loss": 3.4573066234588623, |
| "eval_runtime": 217.3298, |
| "eval_samples_per_second": 555.405, |
| "eval_steps_per_second": 5.788, |
| "step": 90000 |
| }, |
| { |
| "epoch": 3.8148632129157356, |
| "grad_norm": 1.0415303707122803, |
| "learning_rate": 8.619388586373967e-05, |
| "loss": 3.7254, |
| "step": 90500 |
| }, |
| { |
| "epoch": 3.835939805252287, |
| "grad_norm": 1.1311365365982056, |
| "learning_rate": 8.610813360202375e-05, |
| "loss": 3.7163, |
| "step": 91000 |
| }, |
| { |
| "epoch": 3.857016397588838, |
| "grad_norm": 1.0525552034378052, |
| "learning_rate": 8.602238134030785e-05, |
| "loss": 3.7215, |
| "step": 91500 |
| }, |
| { |
| "epoch": 3.878092989925389, |
| "grad_norm": 1.0826270580291748, |
| "learning_rate": 8.593662907859195e-05, |
| "loss": 3.7074, |
| "step": 92000 |
| }, |
| { |
| "epoch": 3.8991695822619397, |
| "grad_norm": 1.0074262619018555, |
| "learning_rate": 8.585087681687605e-05, |
| "loss": 3.7103, |
| "step": 92500 |
| }, |
| { |
| "epoch": 3.920246174598491, |
| "grad_norm": 1.0180414915084839, |
| "learning_rate": 8.576512455516015e-05, |
| "loss": 3.708, |
| "step": 93000 |
| }, |
| { |
| "epoch": 3.941322766935042, |
| "grad_norm": 1.0961127281188965, |
| "learning_rate": 8.567937229344425e-05, |
| "loss": 3.7021, |
| "step": 93500 |
| }, |
| { |
| "epoch": 3.962399359271593, |
| "grad_norm": 1.0143821239471436, |
| "learning_rate": 8.559362003172834e-05, |
| "loss": 3.701, |
| "step": 94000 |
| }, |
| { |
| "epoch": 3.983475951608144, |
| "grad_norm": 1.0110738277435303, |
| "learning_rate": 8.550786777001244e-05, |
| "loss": 3.6968, |
| "step": 94500 |
| }, |
| { |
| "epoch": 4.004552543944695, |
| "grad_norm": 1.1116015911102295, |
| "learning_rate": 8.542211550829654e-05, |
| "loss": 3.6917, |
| "step": 95000 |
| }, |
| { |
| "epoch": 4.025629136281246, |
| "grad_norm": 1.0174391269683838, |
| "learning_rate": 8.533636324658063e-05, |
| "loss": 3.6825, |
| "step": 95500 |
| }, |
| { |
| "epoch": 4.046705728617797, |
| "grad_norm": 1.1200429201126099, |
| "learning_rate": 8.525061098486474e-05, |
| "loss": 3.684, |
| "step": 96000 |
| }, |
| { |
| "epoch": 4.067782320954348, |
| "grad_norm": 1.0700314044952393, |
| "learning_rate": 8.516485872314884e-05, |
| "loss": 3.6783, |
| "step": 96500 |
| }, |
| { |
| "epoch": 4.088858913290899, |
| "grad_norm": 0.9881861805915833, |
| "learning_rate": 8.507910646143292e-05, |
| "loss": 3.676, |
| "step": 97000 |
| }, |
| { |
| "epoch": 4.10993550562745, |
| "grad_norm": 1.1852542161941528, |
| "learning_rate": 8.499335419971702e-05, |
| "loss": 3.6726, |
| "step": 97500 |
| }, |
| { |
| "epoch": 4.1310120979640015, |
| "grad_norm": 1.059182047843933, |
| "learning_rate": 8.490760193800112e-05, |
| "loss": 3.6692, |
| "step": 98000 |
| }, |
| { |
| "epoch": 4.152088690300552, |
| "grad_norm": 1.0494749546051025, |
| "learning_rate": 8.482184967628522e-05, |
| "loss": 3.6671, |
| "step": 98500 |
| }, |
| { |
| "epoch": 4.173165282637103, |
| "grad_norm": 1.0849487781524658, |
| "learning_rate": 8.473609741456931e-05, |
| "loss": 3.6696, |
| "step": 99000 |
| }, |
| { |
| "epoch": 4.194241874973654, |
| "grad_norm": 1.1195166110992432, |
| "learning_rate": 8.46503451528534e-05, |
| "loss": 3.6685, |
| "step": 99500 |
| }, |
| { |
| "epoch": 4.2153184673102055, |
| "grad_norm": 1.0703861713409424, |
| "learning_rate": 8.456459289113751e-05, |
| "loss": 3.6613, |
| "step": 100000 |
| }, |
| { |
| "epoch": 4.2153184673102055, |
| "eval_accuracy": 0.6845051026158379, |
| "eval_loss": 3.404633045196533, |
| "eval_runtime": 209.2542, |
| "eval_samples_per_second": 576.839, |
| "eval_steps_per_second": 6.012, |
| "step": 100000 |
| }, |
| { |
| "epoch": 4.236395059646756, |
| "grad_norm": 1.1069508790969849, |
| "learning_rate": 8.447884062942161e-05, |
| "loss": 3.6633, |
| "step": 100500 |
| }, |
| { |
| "epoch": 4.257471651983307, |
| "grad_norm": 1.0682977437973022, |
| "learning_rate": 8.43930883677057e-05, |
| "loss": 3.6592, |
| "step": 101000 |
| }, |
| { |
| "epoch": 4.278548244319858, |
| "grad_norm": 1.07243812084198, |
| "learning_rate": 8.430733610598979e-05, |
| "loss": 3.6535, |
| "step": 101500 |
| }, |
| { |
| "epoch": 4.29962483665641, |
| "grad_norm": 1.1160458326339722, |
| "learning_rate": 8.42215838442739e-05, |
| "loss": 3.6524, |
| "step": 102000 |
| }, |
| { |
| "epoch": 4.32070142899296, |
| "grad_norm": 1.0621626377105713, |
| "learning_rate": 8.413583158255799e-05, |
| "loss": 3.6496, |
| "step": 102500 |
| }, |
| { |
| "epoch": 4.341778021329511, |
| "grad_norm": 1.0565812587738037, |
| "learning_rate": 8.405007932084209e-05, |
| "loss": 3.6474, |
| "step": 103000 |
| }, |
| { |
| "epoch": 4.362854613666062, |
| "grad_norm": 1.041121482849121, |
| "learning_rate": 8.396432705912619e-05, |
| "loss": 3.6464, |
| "step": 103500 |
| }, |
| { |
| "epoch": 4.383931206002614, |
| "grad_norm": 1.0368785858154297, |
| "learning_rate": 8.387857479741028e-05, |
| "loss": 3.6493, |
| "step": 104000 |
| }, |
| { |
| "epoch": 4.405007798339165, |
| "grad_norm": 1.1039139032363892, |
| "learning_rate": 8.379282253569438e-05, |
| "loss": 3.6421, |
| "step": 104500 |
| }, |
| { |
| "epoch": 4.426084390675715, |
| "grad_norm": 1.0377544164657593, |
| "learning_rate": 8.370707027397848e-05, |
| "loss": 3.6393, |
| "step": 105000 |
| }, |
| { |
| "epoch": 4.4471609830122665, |
| "grad_norm": 1.1220662593841553, |
| "learning_rate": 8.362131801226258e-05, |
| "loss": 3.6413, |
| "step": 105500 |
| }, |
| { |
| "epoch": 4.468237575348818, |
| "grad_norm": 1.0624064207077026, |
| "learning_rate": 8.353556575054668e-05, |
| "loss": 3.6339, |
| "step": 106000 |
| }, |
| { |
| "epoch": 4.489314167685369, |
| "grad_norm": 1.1697174310684204, |
| "learning_rate": 8.344981348883077e-05, |
| "loss": 3.6359, |
| "step": 106500 |
| }, |
| { |
| "epoch": 4.510390760021919, |
| "grad_norm": 1.0498591661453247, |
| "learning_rate": 8.336406122711486e-05, |
| "loss": 3.6303, |
| "step": 107000 |
| }, |
| { |
| "epoch": 4.5314673523584705, |
| "grad_norm": 1.0978573560714722, |
| "learning_rate": 8.327830896539897e-05, |
| "loss": 3.6342, |
| "step": 107500 |
| }, |
| { |
| "epoch": 4.552543944695022, |
| "grad_norm": 1.0736089944839478, |
| "learning_rate": 8.319255670368307e-05, |
| "loss": 3.6305, |
| "step": 108000 |
| }, |
| { |
| "epoch": 4.573620537031573, |
| "grad_norm": 1.100433111190796, |
| "learning_rate": 8.310680444196715e-05, |
| "loss": 3.6243, |
| "step": 108500 |
| }, |
| { |
| "epoch": 4.594697129368123, |
| "grad_norm": 1.0966746807098389, |
| "learning_rate": 8.302105218025125e-05, |
| "loss": 3.6182, |
| "step": 109000 |
| }, |
| { |
| "epoch": 4.615773721704675, |
| "grad_norm": 1.1302976608276367, |
| "learning_rate": 8.293529991853536e-05, |
| "loss": 3.6236, |
| "step": 109500 |
| }, |
| { |
| "epoch": 4.636850314041226, |
| "grad_norm": 1.1187342405319214, |
| "learning_rate": 8.284954765681945e-05, |
| "loss": 3.6198, |
| "step": 110000 |
| }, |
| { |
| "epoch": 4.636850314041226, |
| "eval_accuracy": 0.6909631255102251, |
| "eval_loss": 3.36376953125, |
| "eval_runtime": 213.0816, |
| "eval_samples_per_second": 566.478, |
| "eval_steps_per_second": 5.904, |
| "step": 110000 |
| }, |
| { |
| "epoch": 4.657926906377777, |
| "grad_norm": 1.0469653606414795, |
| "learning_rate": 8.276379539510355e-05, |
| "loss": 3.6163, |
| "step": 110500 |
| }, |
| { |
| "epoch": 4.679003498714328, |
| "grad_norm": 1.0933640003204346, |
| "learning_rate": 8.267804313338765e-05, |
| "loss": 3.6149, |
| "step": 111000 |
| }, |
| { |
| "epoch": 4.700080091050879, |
| "grad_norm": 1.1866936683654785, |
| "learning_rate": 8.259229087167174e-05, |
| "loss": 3.6108, |
| "step": 111500 |
| }, |
| { |
| "epoch": 4.72115668338743, |
| "grad_norm": 1.0143624544143677, |
| "learning_rate": 8.250653860995584e-05, |
| "loss": 3.6108, |
| "step": 112000 |
| }, |
| { |
| "epoch": 4.742233275723981, |
| "grad_norm": 1.2478309869766235, |
| "learning_rate": 8.242078634823994e-05, |
| "loss": 3.6075, |
| "step": 112500 |
| }, |
| { |
| "epoch": 4.763309868060532, |
| "grad_norm": 1.0386154651641846, |
| "learning_rate": 8.233503408652403e-05, |
| "loss": 3.6083, |
| "step": 113000 |
| }, |
| { |
| "epoch": 4.784386460397083, |
| "grad_norm": 1.1464039087295532, |
| "learning_rate": 8.224928182480814e-05, |
| "loss": 3.6073, |
| "step": 113500 |
| }, |
| { |
| "epoch": 4.805463052733634, |
| "grad_norm": 1.0778224468231201, |
| "learning_rate": 8.216352956309224e-05, |
| "loss": 3.5974, |
| "step": 114000 |
| }, |
| { |
| "epoch": 4.826539645070185, |
| "grad_norm": 1.0256168842315674, |
| "learning_rate": 8.207777730137632e-05, |
| "loss": 3.5987, |
| "step": 114500 |
| }, |
| { |
| "epoch": 4.847616237406736, |
| "grad_norm": 1.0351721048355103, |
| "learning_rate": 8.199202503966042e-05, |
| "loss": 3.5976, |
| "step": 115000 |
| }, |
| { |
| "epoch": 4.868692829743287, |
| "grad_norm": 1.1176834106445312, |
| "learning_rate": 8.190627277794453e-05, |
| "loss": 3.5963, |
| "step": 115500 |
| }, |
| { |
| "epoch": 4.889769422079838, |
| "grad_norm": 1.1105574369430542, |
| "learning_rate": 8.182052051622862e-05, |
| "loss": 3.5913, |
| "step": 116000 |
| }, |
| { |
| "epoch": 4.910846014416389, |
| "grad_norm": 1.0621904134750366, |
| "learning_rate": 8.173476825451271e-05, |
| "loss": 3.5928, |
| "step": 116500 |
| }, |
| { |
| "epoch": 4.9319226067529405, |
| "grad_norm": 1.0619655847549438, |
| "learning_rate": 8.164901599279681e-05, |
| "loss": 3.5901, |
| "step": 117000 |
| }, |
| { |
| "epoch": 4.952999199089491, |
| "grad_norm": 1.0421462059020996, |
| "learning_rate": 8.156326373108091e-05, |
| "loss": 3.5915, |
| "step": 117500 |
| }, |
| { |
| "epoch": 4.974075791426042, |
| "grad_norm": 1.0154445171356201, |
| "learning_rate": 8.147751146936501e-05, |
| "loss": 3.5898, |
| "step": 118000 |
| }, |
| { |
| "epoch": 4.995152383762593, |
| "grad_norm": 1.1793843507766724, |
| "learning_rate": 8.139175920764911e-05, |
| "loss": 3.5814, |
| "step": 118500 |
| }, |
| { |
| "epoch": 5.0162289760991445, |
| "grad_norm": 1.061042070388794, |
| "learning_rate": 8.13060069459332e-05, |
| "loss": 3.5815, |
| "step": 119000 |
| }, |
| { |
| "epoch": 5.037305568435695, |
| "grad_norm": 1.0207319259643555, |
| "learning_rate": 8.12202546842173e-05, |
| "loss": 3.576, |
| "step": 119500 |
| }, |
| { |
| "epoch": 5.058382160772246, |
| "grad_norm": 1.0275322198867798, |
| "learning_rate": 8.11345024225014e-05, |
| "loss": 3.5769, |
| "step": 120000 |
| }, |
| { |
| "epoch": 5.058382160772246, |
| "eval_accuracy": 0.6971917734470168, |
| "eval_loss": 3.323504686355591, |
| "eval_runtime": 208.5667, |
| "eval_samples_per_second": 578.741, |
| "eval_steps_per_second": 6.032, |
| "step": 120000 |
| }, |
| { |
| "epoch": 5.079458753108797, |
| "grad_norm": 1.0992586612701416, |
| "learning_rate": 8.104875016078549e-05, |
| "loss": 3.5716, |
| "step": 120500 |
| }, |
| { |
| "epoch": 5.100535345445349, |
| "grad_norm": 1.1391918659210205, |
| "learning_rate": 8.09629978990696e-05, |
| "loss": 3.5696, |
| "step": 121000 |
| }, |
| { |
| "epoch": 5.1216119377819, |
| "grad_norm": 1.0380761623382568, |
| "learning_rate": 8.08772456373537e-05, |
| "loss": 3.571, |
| "step": 121500 |
| }, |
| { |
| "epoch": 5.14268853011845, |
| "grad_norm": 1.1487315893173218, |
| "learning_rate": 8.079149337563778e-05, |
| "loss": 3.5674, |
| "step": 122000 |
| }, |
| { |
| "epoch": 5.163765122455001, |
| "grad_norm": 1.0554592609405518, |
| "learning_rate": 8.070574111392188e-05, |
| "loss": 3.5658, |
| "step": 122500 |
| }, |
| { |
| "epoch": 5.184841714791553, |
| "grad_norm": 1.0251537561416626, |
| "learning_rate": 8.061998885220599e-05, |
| "loss": 3.5647, |
| "step": 123000 |
| }, |
| { |
| "epoch": 5.205918307128104, |
| "grad_norm": 1.0475672483444214, |
| "learning_rate": 8.053423659049008e-05, |
| "loss": 3.5649, |
| "step": 123500 |
| }, |
| { |
| "epoch": 5.226994899464654, |
| "grad_norm": 1.1292567253112793, |
| "learning_rate": 8.044848432877418e-05, |
| "loss": 3.5636, |
| "step": 124000 |
| }, |
| { |
| "epoch": 5.2480714918012055, |
| "grad_norm": 1.1259220838546753, |
| "learning_rate": 8.036273206705827e-05, |
| "loss": 3.5653, |
| "step": 124500 |
| }, |
| { |
| "epoch": 5.269148084137757, |
| "grad_norm": 1.044074535369873, |
| "learning_rate": 8.027697980534237e-05, |
| "loss": 3.5592, |
| "step": 125000 |
| }, |
| { |
| "epoch": 5.290224676474308, |
| "grad_norm": 1.1216226816177368, |
| "learning_rate": 8.019122754362647e-05, |
| "loss": 3.5599, |
| "step": 125500 |
| }, |
| { |
| "epoch": 5.311301268810858, |
| "grad_norm": 1.0395262241363525, |
| "learning_rate": 8.010547528191057e-05, |
| "loss": 3.5578, |
| "step": 126000 |
| }, |
| { |
| "epoch": 5.3323778611474095, |
| "grad_norm": 1.062138319015503, |
| "learning_rate": 8.001972302019465e-05, |
| "loss": 3.5529, |
| "step": 126500 |
| }, |
| { |
| "epoch": 5.353454453483961, |
| "grad_norm": 1.057732105255127, |
| "learning_rate": 7.993397075847876e-05, |
| "loss": 3.5527, |
| "step": 127000 |
| }, |
| { |
| "epoch": 5.374531045820512, |
| "grad_norm": 1.1255929470062256, |
| "learning_rate": 7.984821849676286e-05, |
| "loss": 3.5562, |
| "step": 127500 |
| }, |
| { |
| "epoch": 5.395607638157063, |
| "grad_norm": 1.0545469522476196, |
| "learning_rate": 7.976246623504695e-05, |
| "loss": 3.5514, |
| "step": 128000 |
| }, |
| { |
| "epoch": 5.416684230493614, |
| "grad_norm": 1.0871864557266235, |
| "learning_rate": 7.967671397333105e-05, |
| "loss": 3.5466, |
| "step": 128500 |
| }, |
| { |
| "epoch": 5.437760822830165, |
| "grad_norm": 1.0469675064086914, |
| "learning_rate": 7.959096171161516e-05, |
| "loss": 3.5473, |
| "step": 129000 |
| }, |
| { |
| "epoch": 5.458837415166716, |
| "grad_norm": 1.0500476360321045, |
| "learning_rate": 7.950520944989924e-05, |
| "loss": 3.5455, |
| "step": 129500 |
| }, |
| { |
| "epoch": 5.479914007503266, |
| "grad_norm": 1.0783239603042603, |
| "learning_rate": 7.941945718818334e-05, |
| "loss": 3.5446, |
| "step": 130000 |
| }, |
| { |
| "epoch": 5.479914007503266, |
| "eval_accuracy": 0.7020662781859021, |
| "eval_loss": 3.2929298877716064, |
| "eval_runtime": 208.5101, |
| "eval_samples_per_second": 578.898, |
| "eval_steps_per_second": 6.033, |
| "step": 130000 |
| }, |
| { |
| "epoch": 5.500990599839818, |
| "grad_norm": 1.065730094909668, |
| "learning_rate": 7.933370492646744e-05, |
| "loss": 3.5485, |
| "step": 130500 |
| }, |
| { |
| "epoch": 5.522067192176369, |
| "grad_norm": 1.077804684638977, |
| "learning_rate": 7.924795266475154e-05, |
| "loss": 3.5404, |
| "step": 131000 |
| }, |
| { |
| "epoch": 5.54314378451292, |
| "grad_norm": 1.085261583328247, |
| "learning_rate": 7.916220040303564e-05, |
| "loss": 3.535, |
| "step": 131500 |
| }, |
| { |
| "epoch": 5.564220376849471, |
| "grad_norm": 1.103593349456787, |
| "learning_rate": 7.907644814131972e-05, |
| "loss": 3.5423, |
| "step": 132000 |
| }, |
| { |
| "epoch": 5.585296969186022, |
| "grad_norm": 1.0456616878509521, |
| "learning_rate": 7.899069587960383e-05, |
| "loss": 3.535, |
| "step": 132500 |
| }, |
| { |
| "epoch": 5.606373561522573, |
| "grad_norm": 1.4548872709274292, |
| "learning_rate": 7.890494361788793e-05, |
| "loss": 3.5369, |
| "step": 133000 |
| }, |
| { |
| "epoch": 5.627450153859124, |
| "grad_norm": 1.138386845588684, |
| "learning_rate": 7.881919135617202e-05, |
| "loss": 3.5317, |
| "step": 133500 |
| }, |
| { |
| "epoch": 5.648526746195675, |
| "grad_norm": 1.0792587995529175, |
| "learning_rate": 7.873343909445611e-05, |
| "loss": 3.5286, |
| "step": 134000 |
| }, |
| { |
| "epoch": 5.669603338532226, |
| "grad_norm": 1.0684083700180054, |
| "learning_rate": 7.864768683274023e-05, |
| "loss": 3.538, |
| "step": 134500 |
| }, |
| { |
| "epoch": 5.690679930868777, |
| "grad_norm": 4.212080001831055, |
| "learning_rate": 7.856193457102431e-05, |
| "loss": 3.5348, |
| "step": 135000 |
| }, |
| { |
| "epoch": 5.711756523205328, |
| "grad_norm": 1.0972367525100708, |
| "learning_rate": 7.847618230930841e-05, |
| "loss": 3.5314, |
| "step": 135500 |
| }, |
| { |
| "epoch": 5.7328331155418795, |
| "grad_norm": 9.837974548339844, |
| "learning_rate": 7.839043004759251e-05, |
| "loss": 3.533, |
| "step": 136000 |
| }, |
| { |
| "epoch": 5.75390970787843, |
| "grad_norm": 1.0941179990768433, |
| "learning_rate": 7.83046777858766e-05, |
| "loss": 3.5285, |
| "step": 136500 |
| }, |
| { |
| "epoch": 5.774986300214981, |
| "grad_norm": 7.2343034744262695, |
| "learning_rate": 7.82189255241607e-05, |
| "loss": 3.5265, |
| "step": 137000 |
| }, |
| { |
| "epoch": 5.796062892551532, |
| "grad_norm": 1.14237380027771, |
| "learning_rate": 7.81331732624448e-05, |
| "loss": 3.5237, |
| "step": 137500 |
| }, |
| { |
| "epoch": 5.8171394848880835, |
| "grad_norm": 1.39426851272583, |
| "learning_rate": 7.804742100072889e-05, |
| "loss": 3.5248, |
| "step": 138000 |
| }, |
| { |
| "epoch": 5.838216077224635, |
| "grad_norm": 1.11428964138031, |
| "learning_rate": 7.7961668739013e-05, |
| "loss": 3.5271, |
| "step": 138500 |
| }, |
| { |
| "epoch": 5.859292669561185, |
| "grad_norm": 1.092055082321167, |
| "learning_rate": 7.78759164772971e-05, |
| "loss": 3.5217, |
| "step": 139000 |
| }, |
| { |
| "epoch": 5.880369261897736, |
| "grad_norm": 1.0713292360305786, |
| "learning_rate": 7.779016421558118e-05, |
| "loss": 3.52, |
| "step": 139500 |
| }, |
| { |
| "epoch": 5.901445854234288, |
| "grad_norm": 1.0777440071105957, |
| "learning_rate": 7.770441195386528e-05, |
| "loss": 3.5226, |
| "step": 140000 |
| }, |
| { |
| "epoch": 5.901445854234288, |
| "eval_accuracy": 0.7061789096453397, |
| "eval_loss": 3.2670023441314697, |
| "eval_runtime": 207.8715, |
| "eval_samples_per_second": 580.676, |
| "eval_steps_per_second": 6.052, |
| "step": 140000 |
| }, |
| { |
| "epoch": 5.922522446570838, |
| "grad_norm": 1.1727843284606934, |
| "learning_rate": 7.761865969214939e-05, |
| "loss": 3.5171, |
| "step": 140500 |
| }, |
| { |
| "epoch": 5.943599038907389, |
| "grad_norm": 1.1608026027679443, |
| "learning_rate": 7.753290743043348e-05, |
| "loss": 3.521, |
| "step": 141000 |
| }, |
| { |
| "epoch": 5.96467563124394, |
| "grad_norm": 1.2130433320999146, |
| "learning_rate": 7.744715516871758e-05, |
| "loss": 3.5179, |
| "step": 141500 |
| }, |
| { |
| "epoch": 5.985752223580492, |
| "grad_norm": 1.0992703437805176, |
| "learning_rate": 7.736140290700167e-05, |
| "loss": 3.5136, |
| "step": 142000 |
| }, |
| { |
| "epoch": 6.006828815917043, |
| "grad_norm": 1.1688728332519531, |
| "learning_rate": 7.727565064528577e-05, |
| "loss": 3.5113, |
| "step": 142500 |
| }, |
| { |
| "epoch": 6.027905408253593, |
| "grad_norm": 1.1061811447143555, |
| "learning_rate": 7.718989838356987e-05, |
| "loss": 3.5088, |
| "step": 143000 |
| }, |
| { |
| "epoch": 6.0489820005901445, |
| "grad_norm": 1.0854393243789673, |
| "learning_rate": 7.710414612185397e-05, |
| "loss": 3.5118, |
| "step": 143500 |
| }, |
| { |
| "epoch": 6.070058592926696, |
| "grad_norm": 1.060617446899414, |
| "learning_rate": 7.701839386013807e-05, |
| "loss": 3.5065, |
| "step": 144000 |
| }, |
| { |
| "epoch": 6.091135185263247, |
| "grad_norm": 1.086622953414917, |
| "learning_rate": 7.693264159842217e-05, |
| "loss": 3.5028, |
| "step": 144500 |
| }, |
| { |
| "epoch": 6.112211777599797, |
| "grad_norm": 1.0894832611083984, |
| "learning_rate": 7.684688933670626e-05, |
| "loss": 3.502, |
| "step": 145000 |
| }, |
| { |
| "epoch": 6.1332883699363485, |
| "grad_norm": 1.0915542840957642, |
| "learning_rate": 7.676113707499035e-05, |
| "loss": 3.5029, |
| "step": 145500 |
| }, |
| { |
| "epoch": 6.1543649622729, |
| "grad_norm": 1.0593432188034058, |
| "learning_rate": 7.667538481327446e-05, |
| "loss": 3.5014, |
| "step": 146000 |
| }, |
| { |
| "epoch": 6.175441554609451, |
| "grad_norm": 1.1058335304260254, |
| "learning_rate": 7.658963255155856e-05, |
| "loss": 3.5035, |
| "step": 146500 |
| }, |
| { |
| "epoch": 6.196518146946001, |
| "grad_norm": 1.0519059896469116, |
| "learning_rate": 7.650388028984264e-05, |
| "loss": 3.4978, |
| "step": 147000 |
| }, |
| { |
| "epoch": 6.217594739282553, |
| "grad_norm": 1.0541678667068481, |
| "learning_rate": 7.641812802812674e-05, |
| "loss": 3.4964, |
| "step": 147500 |
| }, |
| { |
| "epoch": 6.238671331619104, |
| "grad_norm": 1.059885859489441, |
| "learning_rate": 7.633237576641084e-05, |
| "loss": 3.4937, |
| "step": 148000 |
| }, |
| { |
| "epoch": 6.259747923955655, |
| "grad_norm": 1.0930180549621582, |
| "learning_rate": 7.624662350469494e-05, |
| "loss": 3.4968, |
| "step": 148500 |
| }, |
| { |
| "epoch": 6.280824516292206, |
| "grad_norm": 1.070291519165039, |
| "learning_rate": 7.616087124297904e-05, |
| "loss": 3.5003, |
| "step": 149000 |
| }, |
| { |
| "epoch": 6.301901108628757, |
| "grad_norm": 1.0394887924194336, |
| "learning_rate": 7.607511898126313e-05, |
| "loss": 3.4946, |
| "step": 149500 |
| }, |
| { |
| "epoch": 6.322977700965308, |
| "grad_norm": 1.049215316772461, |
| "learning_rate": 7.598936671954723e-05, |
| "loss": 3.493, |
| "step": 150000 |
| }, |
| { |
| "epoch": 6.322977700965308, |
| "eval_accuracy": 0.7098668796045495, |
| "eval_loss": 3.246452569961548, |
| "eval_runtime": 207.0943, |
| "eval_samples_per_second": 582.855, |
| "eval_steps_per_second": 6.075, |
| "step": 150000 |
| }, |
| { |
| "epoch": 6.344054293301859, |
| "grad_norm": 1.1335116624832153, |
| "learning_rate": 7.590361445783133e-05, |
| "loss": 3.4923, |
| "step": 150500 |
| }, |
| { |
| "epoch": 6.36513088563841, |
| "grad_norm": 1.0092912912368774, |
| "learning_rate": 7.581786219611543e-05, |
| "loss": 3.4921, |
| "step": 151000 |
| }, |
| { |
| "epoch": 6.386207477974961, |
| "grad_norm": 1.1098837852478027, |
| "learning_rate": 7.573210993439951e-05, |
| "loss": 3.4901, |
| "step": 151500 |
| }, |
| { |
| "epoch": 6.407284070311512, |
| "grad_norm": 1.1028728485107422, |
| "learning_rate": 7.564635767268363e-05, |
| "loss": 3.4888, |
| "step": 152000 |
| }, |
| { |
| "epoch": 6.428360662648063, |
| "grad_norm": 1.0581657886505127, |
| "learning_rate": 7.556060541096772e-05, |
| "loss": 3.49, |
| "step": 152500 |
| }, |
| { |
| "epoch": 6.449437254984614, |
| "grad_norm": 1.0522348880767822, |
| "learning_rate": 7.547485314925181e-05, |
| "loss": 3.4896, |
| "step": 153000 |
| }, |
| { |
| "epoch": 6.470513847321165, |
| "grad_norm": 1.7530624866485596, |
| "learning_rate": 7.538910088753591e-05, |
| "loss": 3.4846, |
| "step": 153500 |
| }, |
| { |
| "epoch": 6.491590439657716, |
| "grad_norm": 1.1856136322021484, |
| "learning_rate": 7.530334862582002e-05, |
| "loss": 3.4839, |
| "step": 154000 |
| }, |
| { |
| "epoch": 6.512667031994267, |
| "grad_norm": 1.0592873096466064, |
| "learning_rate": 7.52175963641041e-05, |
| "loss": 3.4958, |
| "step": 154500 |
| }, |
| { |
| "epoch": 6.5337436243308185, |
| "grad_norm": 1.0945205688476562, |
| "learning_rate": 7.51318441023882e-05, |
| "loss": 3.4901, |
| "step": 155000 |
| }, |
| { |
| "epoch": 6.55482021666737, |
| "grad_norm": 1.0278271436691284, |
| "learning_rate": 7.50460918406723e-05, |
| "loss": 3.4854, |
| "step": 155500 |
| }, |
| { |
| "epoch": 6.57589680900392, |
| "grad_norm": 2.766758918762207, |
| "learning_rate": 7.49603395789564e-05, |
| "loss": 3.4858, |
| "step": 156000 |
| }, |
| { |
| "epoch": 6.596973401340471, |
| "grad_norm": 1.1423465013504028, |
| "learning_rate": 7.48745873172405e-05, |
| "loss": 3.4793, |
| "step": 156500 |
| }, |
| { |
| "epoch": 6.6180499936770225, |
| "grad_norm": 1.095022201538086, |
| "learning_rate": 7.47888350555246e-05, |
| "loss": 3.4774, |
| "step": 157000 |
| }, |
| { |
| "epoch": 6.639126586013573, |
| "grad_norm": 1.1431653499603271, |
| "learning_rate": 7.47030827938087e-05, |
| "loss": 3.4796, |
| "step": 157500 |
| }, |
| { |
| "epoch": 6.660203178350124, |
| "grad_norm": 1.220080852508545, |
| "learning_rate": 7.461733053209279e-05, |
| "loss": 3.4798, |
| "step": 158000 |
| }, |
| { |
| "epoch": 6.681279770686675, |
| "grad_norm": 1.1113580465316772, |
| "learning_rate": 7.453157827037689e-05, |
| "loss": 3.4832, |
| "step": 158500 |
| }, |
| { |
| "epoch": 6.702356363023227, |
| "grad_norm": 1.1077700853347778, |
| "learning_rate": 7.444582600866098e-05, |
| "loss": 3.4747, |
| "step": 159000 |
| }, |
| { |
| "epoch": 6.723432955359778, |
| "grad_norm": 1.1509135961532593, |
| "learning_rate": 7.436007374694507e-05, |
| "loss": 3.4791, |
| "step": 159500 |
| }, |
| { |
| "epoch": 6.744509547696328, |
| "grad_norm": 1.1435599327087402, |
| "learning_rate": 7.427432148522919e-05, |
| "loss": 3.4784, |
| "step": 160000 |
| }, |
| { |
| "epoch": 6.744509547696328, |
| "eval_accuracy": 0.7129784528334937, |
| "eval_loss": 3.2291643619537354, |
| "eval_runtime": 205.557, |
| "eval_samples_per_second": 587.214, |
| "eval_steps_per_second": 6.12, |
| "step": 160000 |
| }, |
| { |
| "epoch": 6.765586140032879, |
| "grad_norm": 1.202584981918335, |
| "learning_rate": 7.418856922351327e-05, |
| "loss": 3.4729, |
| "step": 160500 |
| }, |
| { |
| "epoch": 6.786662732369431, |
| "grad_norm": 1.1698274612426758, |
| "learning_rate": 7.410281696179737e-05, |
| "loss": 3.4716, |
| "step": 161000 |
| }, |
| { |
| "epoch": 6.807739324705982, |
| "grad_norm": 1.1688789129257202, |
| "learning_rate": 7.401706470008147e-05, |
| "loss": 3.468, |
| "step": 161500 |
| }, |
| { |
| "epoch": 6.828815917042532, |
| "grad_norm": 1.0961962938308716, |
| "learning_rate": 7.393131243836557e-05, |
| "loss": 3.4668, |
| "step": 162000 |
| }, |
| { |
| "epoch": 6.8498925093790834, |
| "grad_norm": 1.1141690015792847, |
| "learning_rate": 7.384556017664966e-05, |
| "loss": 3.4647, |
| "step": 162500 |
| }, |
| { |
| "epoch": 6.870969101715635, |
| "grad_norm": 1.0851325988769531, |
| "learning_rate": 7.375980791493376e-05, |
| "loss": 3.4647, |
| "step": 163000 |
| }, |
| { |
| "epoch": 6.892045694052186, |
| "grad_norm": 1.1350326538085938, |
| "learning_rate": 7.367405565321786e-05, |
| "loss": 3.4623, |
| "step": 163500 |
| }, |
| { |
| "epoch": 6.913122286388736, |
| "grad_norm": 1.069841980934143, |
| "learning_rate": 7.358830339150196e-05, |
| "loss": 3.4651, |
| "step": 164000 |
| }, |
| { |
| "epoch": 6.9341988787252875, |
| "grad_norm": 1.1034879684448242, |
| "learning_rate": 7.350255112978606e-05, |
| "loss": 3.4669, |
| "step": 164500 |
| }, |
| { |
| "epoch": 6.955275471061839, |
| "grad_norm": 1.1327505111694336, |
| "learning_rate": 7.341679886807014e-05, |
| "loss": 3.4606, |
| "step": 165000 |
| }, |
| { |
| "epoch": 6.97635206339839, |
| "grad_norm": 1.1569044589996338, |
| "learning_rate": 7.333104660635425e-05, |
| "loss": 3.4603, |
| "step": 165500 |
| }, |
| { |
| "epoch": 6.997428655734941, |
| "grad_norm": 1.0856395959854126, |
| "learning_rate": 7.324529434463835e-05, |
| "loss": 3.4601, |
| "step": 166000 |
| }, |
| { |
| "epoch": 7.018505248071492, |
| "grad_norm": 1.1190896034240723, |
| "learning_rate": 7.315954208292244e-05, |
| "loss": 3.4574, |
| "step": 166500 |
| }, |
| { |
| "epoch": 7.039581840408043, |
| "grad_norm": 1.1264320611953735, |
| "learning_rate": 7.307378982120653e-05, |
| "loss": 3.4516, |
| "step": 167000 |
| }, |
| { |
| "epoch": 7.060658432744594, |
| "grad_norm": 1.0932767391204834, |
| "learning_rate": 7.298803755949063e-05, |
| "loss": 3.4523, |
| "step": 167500 |
| }, |
| { |
| "epoch": 7.081735025081145, |
| "grad_norm": 1.0976347923278809, |
| "learning_rate": 7.290228529777473e-05, |
| "loss": 3.4528, |
| "step": 168000 |
| }, |
| { |
| "epoch": 7.102811617417696, |
| "grad_norm": 1.1163744926452637, |
| "learning_rate": 7.281653303605883e-05, |
| "loss": 3.4539, |
| "step": 168500 |
| }, |
| { |
| "epoch": 7.123888209754247, |
| "grad_norm": 1.1104074716567993, |
| "learning_rate": 7.273078077434293e-05, |
| "loss": 3.4561, |
| "step": 169000 |
| }, |
| { |
| "epoch": 7.144964802090798, |
| "grad_norm": 1.1212096214294434, |
| "learning_rate": 7.264502851262703e-05, |
| "loss": 3.4477, |
| "step": 169500 |
| }, |
| { |
| "epoch": 7.166041394427349, |
| "grad_norm": 1.124900460243225, |
| "learning_rate": 7.255927625091112e-05, |
| "loss": 3.4465, |
| "step": 170000 |
| }, |
| { |
| "epoch": 7.166041394427349, |
| "eval_accuracy": 0.7165369292497207, |
| "eval_loss": 3.2073755264282227, |
| "eval_runtime": 201.9857, |
| "eval_samples_per_second": 597.597, |
| "eval_steps_per_second": 6.228, |
| "step": 170000 |
| }, |
| { |
| "epoch": 7.1871179867639, |
| "grad_norm": 1.132774829864502, |
| "learning_rate": 7.247352398919521e-05, |
| "loss": 3.4495, |
| "step": 170500 |
| }, |
| { |
| "epoch": 7.208194579100451, |
| "grad_norm": 1.0994057655334473, |
| "learning_rate": 7.238777172747931e-05, |
| "loss": 3.4488, |
| "step": 171000 |
| }, |
| { |
| "epoch": 7.229271171437002, |
| "grad_norm": 1.0987988710403442, |
| "learning_rate": 7.230201946576342e-05, |
| "loss": 3.4482, |
| "step": 171500 |
| }, |
| { |
| "epoch": 7.250347763773553, |
| "grad_norm": 1.0861831903457642, |
| "learning_rate": 7.22162672040475e-05, |
| "loss": 3.4475, |
| "step": 172000 |
| }, |
| { |
| "epoch": 7.271424356110104, |
| "grad_norm": 1.0958621501922607, |
| "learning_rate": 7.21305149423316e-05, |
| "loss": 3.4445, |
| "step": 172500 |
| }, |
| { |
| "epoch": 7.292500948446655, |
| "grad_norm": 1.1067219972610474, |
| "learning_rate": 7.20447626806157e-05, |
| "loss": 3.4412, |
| "step": 173000 |
| }, |
| { |
| "epoch": 7.313577540783206, |
| "grad_norm": 1.1006122827529907, |
| "learning_rate": 7.19590104188998e-05, |
| "loss": 3.4404, |
| "step": 173500 |
| }, |
| { |
| "epoch": 7.3346541331197574, |
| "grad_norm": 1.094212293624878, |
| "learning_rate": 7.18732581571839e-05, |
| "loss": 3.4433, |
| "step": 174000 |
| }, |
| { |
| "epoch": 7.355730725456308, |
| "grad_norm": 1.1098824739456177, |
| "learning_rate": 7.1787505895468e-05, |
| "loss": 3.4387, |
| "step": 174500 |
| }, |
| { |
| "epoch": 7.376807317792859, |
| "grad_norm": 1.0986790657043457, |
| "learning_rate": 7.17017536337521e-05, |
| "loss": 3.4413, |
| "step": 175000 |
| }, |
| { |
| "epoch": 7.39788391012941, |
| "grad_norm": 1.1846095323562622, |
| "learning_rate": 7.161600137203619e-05, |
| "loss": 3.4423, |
| "step": 175500 |
| }, |
| { |
| "epoch": 7.4189605024659615, |
| "grad_norm": 1.1713793277740479, |
| "learning_rate": 7.153024911032029e-05, |
| "loss": 3.4367, |
| "step": 176000 |
| }, |
| { |
| "epoch": 7.440037094802513, |
| "grad_norm": 1.1812632083892822, |
| "learning_rate": 7.144449684860438e-05, |
| "loss": 3.4392, |
| "step": 176500 |
| }, |
| { |
| "epoch": 7.461113687139063, |
| "grad_norm": 1.0777934789657593, |
| "learning_rate": 7.135874458688849e-05, |
| "loss": 3.44, |
| "step": 177000 |
| }, |
| { |
| "epoch": 7.482190279475614, |
| "grad_norm": 1.1836484670639038, |
| "learning_rate": 7.127299232517259e-05, |
| "loss": 3.4399, |
| "step": 177500 |
| }, |
| { |
| "epoch": 7.503266871812166, |
| "grad_norm": 1.083646535873413, |
| "learning_rate": 7.118724006345667e-05, |
| "loss": 3.4372, |
| "step": 178000 |
| }, |
| { |
| "epoch": 7.524343464148717, |
| "grad_norm": 1.1568636894226074, |
| "learning_rate": 7.110148780174077e-05, |
| "loss": 3.4329, |
| "step": 178500 |
| }, |
| { |
| "epoch": 7.545420056485267, |
| "grad_norm": 1.1540913581848145, |
| "learning_rate": 7.101573554002488e-05, |
| "loss": 3.4329, |
| "step": 179000 |
| }, |
| { |
| "epoch": 7.566496648821818, |
| "grad_norm": 1.0948430299758911, |
| "learning_rate": 7.092998327830897e-05, |
| "loss": 3.4343, |
| "step": 179500 |
| }, |
| { |
| "epoch": 7.58757324115837, |
| "grad_norm": 1.1346430778503418, |
| "learning_rate": 7.084423101659306e-05, |
| "loss": 3.433, |
| "step": 180000 |
| }, |
| { |
| "epoch": 7.58757324115837, |
| "eval_accuracy": 0.7186242811855383, |
| "eval_loss": 3.195988416671753, |
| "eval_runtime": 211.7801, |
| "eval_samples_per_second": 569.959, |
| "eval_steps_per_second": 5.94, |
| "step": 180000 |
| }, |
| { |
| "epoch": 7.608649833494921, |
| "grad_norm": 1.06849205493927, |
| "learning_rate": 7.075847875487716e-05, |
| "loss": 3.4308, |
| "step": 180500 |
| }, |
| { |
| "epoch": 7.629726425831471, |
| "grad_norm": 1.1338483095169067, |
| "learning_rate": 7.067272649316126e-05, |
| "loss": 3.4326, |
| "step": 181000 |
| }, |
| { |
| "epoch": 7.650803018168022, |
| "grad_norm": 1.1990658044815063, |
| "learning_rate": 7.058697423144536e-05, |
| "loss": 3.4286, |
| "step": 181500 |
| }, |
| { |
| "epoch": 7.671879610504574, |
| "grad_norm": 1.111740231513977, |
| "learning_rate": 7.050122196972946e-05, |
| "loss": 3.4268, |
| "step": 182000 |
| }, |
| { |
| "epoch": 7.692956202841125, |
| "grad_norm": 1.1153830289840698, |
| "learning_rate": 7.041546970801354e-05, |
| "loss": 3.429, |
| "step": 182500 |
| }, |
| { |
| "epoch": 7.714032795177676, |
| "grad_norm": 1.0963406562805176, |
| "learning_rate": 7.032971744629765e-05, |
| "loss": 3.4278, |
| "step": 183000 |
| }, |
| { |
| "epoch": 7.7351093875142265, |
| "grad_norm": 1.1566215753555298, |
| "learning_rate": 7.024396518458175e-05, |
| "loss": 3.4315, |
| "step": 183500 |
| }, |
| { |
| "epoch": 7.756185979850778, |
| "grad_norm": 1.0449819564819336, |
| "learning_rate": 7.015821292286584e-05, |
| "loss": 3.4274, |
| "step": 184000 |
| }, |
| { |
| "epoch": 7.777262572187329, |
| "grad_norm": 1.0871033668518066, |
| "learning_rate": 7.007246066114993e-05, |
| "loss": 3.4259, |
| "step": 184500 |
| }, |
| { |
| "epoch": 7.798339164523879, |
| "grad_norm": 1.1184698343276978, |
| "learning_rate": 6.998670839943405e-05, |
| "loss": 3.4248, |
| "step": 185000 |
| }, |
| { |
| "epoch": 7.8194157568604306, |
| "grad_norm": 1.1108282804489136, |
| "learning_rate": 6.990095613771813e-05, |
| "loss": 3.4238, |
| "step": 185500 |
| }, |
| { |
| "epoch": 7.840492349196982, |
| "grad_norm": 1.194790005683899, |
| "learning_rate": 6.981520387600223e-05, |
| "loss": 3.4236, |
| "step": 186000 |
| }, |
| { |
| "epoch": 7.861568941533533, |
| "grad_norm": 1.0745289325714111, |
| "learning_rate": 6.972945161428633e-05, |
| "loss": 3.423, |
| "step": 186500 |
| }, |
| { |
| "epoch": 7.882645533870084, |
| "grad_norm": 1.0792198181152344, |
| "learning_rate": 6.964369935257043e-05, |
| "loss": 3.4224, |
| "step": 187000 |
| }, |
| { |
| "epoch": 7.903722126206635, |
| "grad_norm": 1.1024041175842285, |
| "learning_rate": 6.955794709085452e-05, |
| "loss": 3.4239, |
| "step": 187500 |
| }, |
| { |
| "epoch": 7.924798718543186, |
| "grad_norm": 1.0632470846176147, |
| "learning_rate": 6.947219482913862e-05, |
| "loss": 3.4124, |
| "step": 188000 |
| }, |
| { |
| "epoch": 7.945875310879737, |
| "grad_norm": 1.1112148761749268, |
| "learning_rate": 6.938644256742272e-05, |
| "loss": 3.4218, |
| "step": 188500 |
| }, |
| { |
| "epoch": 7.966951903216288, |
| "grad_norm": 1.105888843536377, |
| "learning_rate": 6.930069030570682e-05, |
| "loss": 3.419, |
| "step": 189000 |
| }, |
| { |
| "epoch": 7.988028495552839, |
| "grad_norm": 1.1686981916427612, |
| "learning_rate": 6.921493804399092e-05, |
| "loss": 3.4161, |
| "step": 189500 |
| }, |
| { |
| "epoch": 8.00910508788939, |
| "grad_norm": 1.124916911125183, |
| "learning_rate": 6.9129185782275e-05, |
| "loss": 3.4171, |
| "step": 190000 |
| }, |
| { |
| "epoch": 8.00910508788939, |
| "eval_accuracy": 0.7212376533274212, |
| "eval_loss": 3.1785213947296143, |
| "eval_runtime": 206.2262, |
| "eval_samples_per_second": 585.309, |
| "eval_steps_per_second": 6.1, |
| "step": 190000 |
| }, |
| { |
| "epoch": 8.03018168022594, |
| "grad_norm": 1.1727031469345093, |
| "learning_rate": 6.904343352055911e-05, |
| "loss": 3.4109, |
| "step": 190500 |
| }, |
| { |
| "epoch": 8.051258272562492, |
| "grad_norm": 1.134281039237976, |
| "learning_rate": 6.895768125884321e-05, |
| "loss": 3.4139, |
| "step": 191000 |
| }, |
| { |
| "epoch": 8.072334864899043, |
| "grad_norm": 1.1211605072021484, |
| "learning_rate": 6.88719289971273e-05, |
| "loss": 3.411, |
| "step": 191500 |
| }, |
| { |
| "epoch": 8.093411457235595, |
| "grad_norm": 1.2221907377243042, |
| "learning_rate": 6.87861767354114e-05, |
| "loss": 3.4059, |
| "step": 192000 |
| }, |
| { |
| "epoch": 8.114488049572145, |
| "grad_norm": 1.1225556135177612, |
| "learning_rate": 6.870042447369551e-05, |
| "loss": 3.4125, |
| "step": 192500 |
| }, |
| { |
| "epoch": 8.135564641908696, |
| "grad_norm": 1.1474628448486328, |
| "learning_rate": 6.861467221197959e-05, |
| "loss": 3.4099, |
| "step": 193000 |
| }, |
| { |
| "epoch": 8.156641234245248, |
| "grad_norm": 1.138868808746338, |
| "learning_rate": 6.852891995026369e-05, |
| "loss": 3.4101, |
| "step": 193500 |
| }, |
| { |
| "epoch": 8.177717826581798, |
| "grad_norm": 1.133005976676941, |
| "learning_rate": 6.844316768854779e-05, |
| "loss": 3.4113, |
| "step": 194000 |
| }, |
| { |
| "epoch": 8.19879441891835, |
| "grad_norm": 1.1481449604034424, |
| "learning_rate": 6.835741542683189e-05, |
| "loss": 3.4095, |
| "step": 194500 |
| }, |
| { |
| "epoch": 8.2198710112549, |
| "grad_norm": 1.0802927017211914, |
| "learning_rate": 6.827166316511599e-05, |
| "loss": 3.4044, |
| "step": 195000 |
| }, |
| { |
| "epoch": 8.24094760359145, |
| "grad_norm": 1.0988690853118896, |
| "learning_rate": 6.818591090340008e-05, |
| "loss": 3.4082, |
| "step": 195500 |
| }, |
| { |
| "epoch": 8.262024195928003, |
| "grad_norm": 1.0715197324752808, |
| "learning_rate": 6.810015864168417e-05, |
| "loss": 3.4033, |
| "step": 196000 |
| }, |
| { |
| "epoch": 8.283100788264553, |
| "grad_norm": 1.0560503005981445, |
| "learning_rate": 6.801440637996828e-05, |
| "loss": 3.4033, |
| "step": 196500 |
| }, |
| { |
| "epoch": 8.304177380601104, |
| "grad_norm": 1.095314860343933, |
| "learning_rate": 6.792865411825238e-05, |
| "loss": 3.4059, |
| "step": 197000 |
| }, |
| { |
| "epoch": 8.325253972937656, |
| "grad_norm": 1.1141575574874878, |
| "learning_rate": 6.784290185653646e-05, |
| "loss": 3.4022, |
| "step": 197500 |
| }, |
| { |
| "epoch": 8.346330565274206, |
| "grad_norm": 1.1695075035095215, |
| "learning_rate": 6.775714959482056e-05, |
| "loss": 3.4023, |
| "step": 198000 |
| }, |
| { |
| "epoch": 8.367407157610758, |
| "grad_norm": 1.0828735828399658, |
| "learning_rate": 6.767139733310467e-05, |
| "loss": 3.4025, |
| "step": 198500 |
| }, |
| { |
| "epoch": 8.388483749947309, |
| "grad_norm": 1.1050857305526733, |
| "learning_rate": 6.758564507138876e-05, |
| "loss": 3.4059, |
| "step": 199000 |
| }, |
| { |
| "epoch": 8.409560342283859, |
| "grad_norm": 1.1212519407272339, |
| "learning_rate": 6.749989280967286e-05, |
| "loss": 3.4056, |
| "step": 199500 |
| }, |
| { |
| "epoch": 8.430636934620411, |
| "grad_norm": 1.086789846420288, |
| "learning_rate": 6.741414054795696e-05, |
| "loss": 3.4016, |
| "step": 200000 |
| }, |
| { |
| "epoch": 8.430636934620411, |
| "eval_accuracy": 0.7236758755008075, |
| "eval_loss": 3.163999557495117, |
| "eval_runtime": 203.7817, |
| "eval_samples_per_second": 592.33, |
| "eval_steps_per_second": 6.173, |
| "step": 200000 |
| }, |
| { |
| "epoch": 8.451713526956961, |
| "grad_norm": 1.1242396831512451, |
| "learning_rate": 6.732838828624105e-05, |
| "loss": 3.4006, |
| "step": 200500 |
| }, |
| { |
| "epoch": 8.472790119293512, |
| "grad_norm": 1.1468968391418457, |
| "learning_rate": 6.724263602452515e-05, |
| "loss": 3.4017, |
| "step": 201000 |
| }, |
| { |
| "epoch": 8.493866711630064, |
| "grad_norm": 1.160112738609314, |
| "learning_rate": 6.715688376280924e-05, |
| "loss": 3.3989, |
| "step": 201500 |
| }, |
| { |
| "epoch": 8.514943303966614, |
| "grad_norm": 1.1590039730072021, |
| "learning_rate": 6.707113150109335e-05, |
| "loss": 3.3992, |
| "step": 202000 |
| }, |
| { |
| "epoch": 8.536019896303166, |
| "grad_norm": 1.119093418121338, |
| "learning_rate": 6.698537923937745e-05, |
| "loss": 3.4028, |
| "step": 202500 |
| }, |
| { |
| "epoch": 8.557096488639717, |
| "grad_norm": 1.1281869411468506, |
| "learning_rate": 6.689962697766153e-05, |
| "loss": 3.3979, |
| "step": 203000 |
| }, |
| { |
| "epoch": 8.578173080976267, |
| "grad_norm": 1.1275852918624878, |
| "learning_rate": 6.681387471594563e-05, |
| "loss": 3.3978, |
| "step": 203500 |
| }, |
| { |
| "epoch": 8.59924967331282, |
| "grad_norm": 1.1040563583374023, |
| "learning_rate": 6.672812245422974e-05, |
| "loss": 3.3961, |
| "step": 204000 |
| }, |
| { |
| "epoch": 8.62032626564937, |
| "grad_norm": 1.1196154356002808, |
| "learning_rate": 6.664237019251383e-05, |
| "loss": 3.3957, |
| "step": 204500 |
| }, |
| { |
| "epoch": 8.64140285798592, |
| "grad_norm": 1.1568843126296997, |
| "learning_rate": 6.655661793079792e-05, |
| "loss": 3.3961, |
| "step": 205000 |
| }, |
| { |
| "epoch": 8.662479450322472, |
| "grad_norm": 1.1717989444732666, |
| "learning_rate": 6.647086566908202e-05, |
| "loss": 3.3911, |
| "step": 205500 |
| }, |
| { |
| "epoch": 8.683556042659022, |
| "grad_norm": 1.1160415410995483, |
| "learning_rate": 6.638511340736612e-05, |
| "loss": 3.3913, |
| "step": 206000 |
| }, |
| { |
| "epoch": 8.704632634995574, |
| "grad_norm": 1.109422206878662, |
| "learning_rate": 6.629936114565022e-05, |
| "loss": 3.3892, |
| "step": 206500 |
| }, |
| { |
| "epoch": 8.725709227332125, |
| "grad_norm": 1.0946357250213623, |
| "learning_rate": 6.621360888393432e-05, |
| "loss": 3.3897, |
| "step": 207000 |
| }, |
| { |
| "epoch": 8.746785819668675, |
| "grad_norm": 1.0983866453170776, |
| "learning_rate": 6.61278566222184e-05, |
| "loss": 3.3833, |
| "step": 207500 |
| }, |
| { |
| "epoch": 8.767862412005227, |
| "grad_norm": 1.0923513174057007, |
| "learning_rate": 6.604210436050251e-05, |
| "loss": 3.39, |
| "step": 208000 |
| }, |
| { |
| "epoch": 8.788939004341778, |
| "grad_norm": 1.1108269691467285, |
| "learning_rate": 6.595635209878661e-05, |
| "loss": 3.3885, |
| "step": 208500 |
| }, |
| { |
| "epoch": 8.81001559667833, |
| "grad_norm": 1.100511074066162, |
| "learning_rate": 6.58705998370707e-05, |
| "loss": 3.3887, |
| "step": 209000 |
| }, |
| { |
| "epoch": 8.83109218901488, |
| "grad_norm": 1.117974877357483, |
| "learning_rate": 6.57848475753548e-05, |
| "loss": 3.3884, |
| "step": 209500 |
| }, |
| { |
| "epoch": 8.85216878135143, |
| "grad_norm": 1.0981616973876953, |
| "learning_rate": 6.569909531363891e-05, |
| "loss": 3.3853, |
| "step": 210000 |
| }, |
| { |
| "epoch": 8.85216878135143, |
| "eval_accuracy": 0.7256641421630827, |
| "eval_loss": 3.153757333755493, |
| "eval_runtime": 207.0906, |
| "eval_samples_per_second": 582.866, |
| "eval_steps_per_second": 6.075, |
| "step": 210000 |
| }, |
| { |
| "epoch": 8.873245373687983, |
| "grad_norm": 1.1012574434280396, |
| "learning_rate": 6.561334305192299e-05, |
| "loss": 3.3871, |
| "step": 210500 |
| }, |
| { |
| "epoch": 8.894321966024533, |
| "grad_norm": 1.1736830472946167, |
| "learning_rate": 6.552759079020709e-05, |
| "loss": 3.3888, |
| "step": 211000 |
| }, |
| { |
| "epoch": 8.915398558361083, |
| "grad_norm": 1.135558843612671, |
| "learning_rate": 6.544183852849119e-05, |
| "loss": 3.3839, |
| "step": 211500 |
| }, |
| { |
| "epoch": 8.936475150697635, |
| "grad_norm": 1.0844402313232422, |
| "learning_rate": 6.535608626677529e-05, |
| "loss": 3.3894, |
| "step": 212000 |
| }, |
| { |
| "epoch": 8.957551743034186, |
| "grad_norm": 1.1128517389297485, |
| "learning_rate": 6.527033400505939e-05, |
| "loss": 3.3872, |
| "step": 212500 |
| }, |
| { |
| "epoch": 8.978628335370738, |
| "grad_norm": 1.102283000946045, |
| "learning_rate": 6.518458174334348e-05, |
| "loss": 3.388, |
| "step": 213000 |
| }, |
| { |
| "epoch": 8.999704927707288, |
| "grad_norm": 1.133675456047058, |
| "learning_rate": 6.509882948162758e-05, |
| "loss": 3.3864, |
| "step": 213500 |
| }, |
| { |
| "epoch": 9.020781520043839, |
| "grad_norm": 1.157792568206787, |
| "learning_rate": 6.501307721991168e-05, |
| "loss": 3.3794, |
| "step": 214000 |
| }, |
| { |
| "epoch": 9.04185811238039, |
| "grad_norm": 1.1257504224777222, |
| "learning_rate": 6.492732495819578e-05, |
| "loss": 3.3763, |
| "step": 214500 |
| }, |
| { |
| "epoch": 9.062934704716941, |
| "grad_norm": 1.1369588375091553, |
| "learning_rate": 6.484157269647986e-05, |
| "loss": 3.3811, |
| "step": 215000 |
| }, |
| { |
| "epoch": 9.084011297053493, |
| "grad_norm": 1.1264878511428833, |
| "learning_rate": 6.475582043476398e-05, |
| "loss": 3.3788, |
| "step": 215500 |
| }, |
| { |
| "epoch": 9.105087889390044, |
| "grad_norm": 1.1348992586135864, |
| "learning_rate": 6.467006817304807e-05, |
| "loss": 3.3761, |
| "step": 216000 |
| }, |
| { |
| "epoch": 9.126164481726594, |
| "grad_norm": 1.124710202217102, |
| "learning_rate": 6.458431591133216e-05, |
| "loss": 3.3705, |
| "step": 216500 |
| }, |
| { |
| "epoch": 9.147241074063146, |
| "grad_norm": 1.1406720876693726, |
| "learning_rate": 6.449856364961626e-05, |
| "loss": 3.3751, |
| "step": 217000 |
| }, |
| { |
| "epoch": 9.168317666399696, |
| "grad_norm": 1.1368287801742554, |
| "learning_rate": 6.441281138790037e-05, |
| "loss": 3.3733, |
| "step": 217500 |
| }, |
| { |
| "epoch": 9.189394258736247, |
| "grad_norm": 1.089103102684021, |
| "learning_rate": 6.432705912618445e-05, |
| "loss": 3.3749, |
| "step": 218000 |
| }, |
| { |
| "epoch": 9.210470851072799, |
| "grad_norm": 1.0838497877120972, |
| "learning_rate": 6.424130686446855e-05, |
| "loss": 3.3765, |
| "step": 218500 |
| }, |
| { |
| "epoch": 9.23154744340935, |
| "grad_norm": 1.1491973400115967, |
| "learning_rate": 6.415555460275265e-05, |
| "loss": 3.3737, |
| "step": 219000 |
| }, |
| { |
| "epoch": 9.252624035745901, |
| "grad_norm": 1.1065986156463623, |
| "learning_rate": 6.406980234103675e-05, |
| "loss": 3.3736, |
| "step": 219500 |
| }, |
| { |
| "epoch": 9.273700628082452, |
| "grad_norm": 1.1861274242401123, |
| "learning_rate": 6.398405007932085e-05, |
| "loss": 3.3776, |
| "step": 220000 |
| }, |
| { |
| "epoch": 9.273700628082452, |
| "eval_accuracy": 0.7278431975143994, |
| "eval_loss": 3.1403961181640625, |
| "eval_runtime": 204.3372, |
| "eval_samples_per_second": 590.72, |
| "eval_steps_per_second": 6.156, |
| "step": 220000 |
| }, |
| { |
| "epoch": 9.294777220419002, |
| "grad_norm": 1.092643141746521, |
| "learning_rate": 6.389829781760495e-05, |
| "loss": 3.3733, |
| "step": 220500 |
| }, |
| { |
| "epoch": 9.315853812755554, |
| "grad_norm": 1.1328953504562378, |
| "learning_rate": 6.381254555588903e-05, |
| "loss": 3.3746, |
| "step": 221000 |
| }, |
| { |
| "epoch": 9.336930405092104, |
| "grad_norm": 1.0782948732376099, |
| "learning_rate": 6.372679329417314e-05, |
| "loss": 3.3698, |
| "step": 221500 |
| }, |
| { |
| "epoch": 9.358006997428657, |
| "grad_norm": 1.1517688035964966, |
| "learning_rate": 6.364104103245724e-05, |
| "loss": 3.3676, |
| "step": 222000 |
| }, |
| { |
| "epoch": 9.379083589765207, |
| "grad_norm": 1.1444026231765747, |
| "learning_rate": 6.355528877074132e-05, |
| "loss": 3.3723, |
| "step": 222500 |
| }, |
| { |
| "epoch": 9.400160182101757, |
| "grad_norm": 1.1651170253753662, |
| "learning_rate": 6.346953650902542e-05, |
| "loss": 3.3709, |
| "step": 223000 |
| }, |
| { |
| "epoch": 9.42123677443831, |
| "grad_norm": 1.139785885810852, |
| "learning_rate": 6.338378424730954e-05, |
| "loss": 3.3703, |
| "step": 223500 |
| }, |
| { |
| "epoch": 9.44231336677486, |
| "grad_norm": 1.1319204568862915, |
| "learning_rate": 6.329803198559362e-05, |
| "loss": 3.3666, |
| "step": 224000 |
| }, |
| { |
| "epoch": 9.46338995911141, |
| "grad_norm": 1.1481612920761108, |
| "learning_rate": 6.321227972387772e-05, |
| "loss": 3.3678, |
| "step": 224500 |
| }, |
| { |
| "epoch": 9.484466551447962, |
| "grad_norm": 1.0957386493682861, |
| "learning_rate": 6.312652746216182e-05, |
| "loss": 3.3634, |
| "step": 225000 |
| }, |
| { |
| "epoch": 9.505543143784513, |
| "grad_norm": 1.0830187797546387, |
| "learning_rate": 6.304077520044591e-05, |
| "loss": 3.3722, |
| "step": 225500 |
| }, |
| { |
| "epoch": 9.526619736121065, |
| "grad_norm": 1.1297203302383423, |
| "learning_rate": 6.295502293873001e-05, |
| "loss": 3.3667, |
| "step": 226000 |
| }, |
| { |
| "epoch": 9.547696328457615, |
| "grad_norm": 1.1397178173065186, |
| "learning_rate": 6.286927067701411e-05, |
| "loss": 3.362, |
| "step": 226500 |
| }, |
| { |
| "epoch": 9.568772920794165, |
| "grad_norm": 1.1798996925354004, |
| "learning_rate": 6.278351841529821e-05, |
| "loss": 3.3672, |
| "step": 227000 |
| }, |
| { |
| "epoch": 9.589849513130718, |
| "grad_norm": 1.1377357244491577, |
| "learning_rate": 6.269776615358231e-05, |
| "loss": 3.3652, |
| "step": 227500 |
| }, |
| { |
| "epoch": 9.610926105467268, |
| "grad_norm": 1.0990251302719116, |
| "learning_rate": 6.26120138918664e-05, |
| "loss": 3.3633, |
| "step": 228000 |
| }, |
| { |
| "epoch": 9.63200269780382, |
| "grad_norm": 1.0902920961380005, |
| "learning_rate": 6.252626163015049e-05, |
| "loss": 3.3636, |
| "step": 228500 |
| }, |
| { |
| "epoch": 9.65307929014037, |
| "grad_norm": 1.1542810201644897, |
| "learning_rate": 6.24405093684346e-05, |
| "loss": 3.3611, |
| "step": 229000 |
| }, |
| { |
| "epoch": 9.67415588247692, |
| "grad_norm": 1.1210051774978638, |
| "learning_rate": 6.23547571067187e-05, |
| "loss": 3.3619, |
| "step": 229500 |
| }, |
| { |
| "epoch": 9.695232474813473, |
| "grad_norm": 1.156883955001831, |
| "learning_rate": 6.226900484500279e-05, |
| "loss": 3.3568, |
| "step": 230000 |
| }, |
| { |
| "epoch": 9.695232474813473, |
| "eval_accuracy": 0.7294716004472328, |
| "eval_loss": 3.1291487216949463, |
| "eval_runtime": 205.5324, |
| "eval_samples_per_second": 587.285, |
| "eval_steps_per_second": 6.121, |
| "step": 230000 |
| }, |
| { |
| "epoch": 9.716309067150023, |
| "grad_norm": 1.1882915496826172, |
| "learning_rate": 6.218325258328688e-05, |
| "loss": 3.3642, |
| "step": 230500 |
| }, |
| { |
| "epoch": 9.737385659486574, |
| "grad_norm": 1.1436213254928589, |
| "learning_rate": 6.209750032157098e-05, |
| "loss": 3.3636, |
| "step": 231000 |
| }, |
| { |
| "epoch": 9.758462251823126, |
| "grad_norm": 1.148887038230896, |
| "learning_rate": 6.201174805985508e-05, |
| "loss": 3.3576, |
| "step": 231500 |
| }, |
| { |
| "epoch": 9.779538844159676, |
| "grad_norm": 1.1576124429702759, |
| "learning_rate": 6.192599579813918e-05, |
| "loss": 3.3594, |
| "step": 232000 |
| }, |
| { |
| "epoch": 9.800615436496226, |
| "grad_norm": 1.124903917312622, |
| "learning_rate": 6.184024353642328e-05, |
| "loss": 3.3578, |
| "step": 232500 |
| }, |
| { |
| "epoch": 9.821692028832778, |
| "grad_norm": 1.1633460521697998, |
| "learning_rate": 6.175449127470738e-05, |
| "loss": 3.3566, |
| "step": 233000 |
| }, |
| { |
| "epoch": 9.842768621169329, |
| "grad_norm": 1.1234873533248901, |
| "learning_rate": 6.166873901299147e-05, |
| "loss": 3.361, |
| "step": 233500 |
| }, |
| { |
| "epoch": 9.863845213505881, |
| "grad_norm": 1.1294035911560059, |
| "learning_rate": 6.158298675127557e-05, |
| "loss": 3.3562, |
| "step": 234000 |
| }, |
| { |
| "epoch": 9.884921805842431, |
| "grad_norm": 1.1272541284561157, |
| "learning_rate": 6.149723448955966e-05, |
| "loss": 3.3599, |
| "step": 234500 |
| }, |
| { |
| "epoch": 9.905998398178982, |
| "grad_norm": 1.1447759866714478, |
| "learning_rate": 6.141148222784377e-05, |
| "loss": 3.3555, |
| "step": 235000 |
| }, |
| { |
| "epoch": 9.927074990515534, |
| "grad_norm": 1.1419342756271362, |
| "learning_rate": 6.132572996612787e-05, |
| "loss": 3.3536, |
| "step": 235500 |
| }, |
| { |
| "epoch": 9.948151582852084, |
| "grad_norm": 1.1533398628234863, |
| "learning_rate": 6.123997770441195e-05, |
| "loss": 3.3578, |
| "step": 236000 |
| }, |
| { |
| "epoch": 9.969228175188636, |
| "grad_norm": 1.144256830215454, |
| "learning_rate": 6.115422544269605e-05, |
| "loss": 3.3585, |
| "step": 236500 |
| }, |
| { |
| "epoch": 9.990304767525187, |
| "grad_norm": 1.1676476001739502, |
| "learning_rate": 6.106847318098015e-05, |
| "loss": 3.3565, |
| "step": 237000 |
| }, |
| { |
| "epoch": 10.011381359861737, |
| "grad_norm": 1.108898401260376, |
| "learning_rate": 6.098272091926425e-05, |
| "loss": 3.3542, |
| "step": 237500 |
| }, |
| { |
| "epoch": 10.032457952198289, |
| "grad_norm": 1.1816314458847046, |
| "learning_rate": 6.0896968657548345e-05, |
| "loss": 3.3531, |
| "step": 238000 |
| }, |
| { |
| "epoch": 10.05353454453484, |
| "grad_norm": 1.114430546760559, |
| "learning_rate": 6.081121639583244e-05, |
| "loss": 3.3504, |
| "step": 238500 |
| }, |
| { |
| "epoch": 10.07461113687139, |
| "grad_norm": 1.098649263381958, |
| "learning_rate": 6.072546413411654e-05, |
| "loss": 3.3494, |
| "step": 239000 |
| }, |
| { |
| "epoch": 10.095687729207942, |
| "grad_norm": 1.1124765872955322, |
| "learning_rate": 6.063971187240064e-05, |
| "loss": 3.3504, |
| "step": 239500 |
| }, |
| { |
| "epoch": 10.116764321544492, |
| "grad_norm": 1.1391160488128662, |
| "learning_rate": 6.055395961068473e-05, |
| "loss": 3.3441, |
| "step": 240000 |
| }, |
| { |
| "epoch": 10.116764321544492, |
| "eval_accuracy": 0.7314897905026054, |
| "eval_loss": 3.120478391647339, |
| "eval_runtime": 212.8499, |
| "eval_samples_per_second": 567.094, |
| "eval_steps_per_second": 5.91, |
| "step": 240000 |
| }, |
| { |
| "epoch": 10.137840913881044, |
| "grad_norm": 1.2106882333755493, |
| "learning_rate": 6.046820734896883e-05, |
| "loss": 3.3461, |
| "step": 240500 |
| }, |
| { |
| "epoch": 10.158917506217595, |
| "grad_norm": 1.1477400064468384, |
| "learning_rate": 6.038245508725293e-05, |
| "loss": 3.348, |
| "step": 241000 |
| }, |
| { |
| "epoch": 10.179994098554145, |
| "grad_norm": 1.1499747037887573, |
| "learning_rate": 6.029670282553702e-05, |
| "loss": 3.3464, |
| "step": 241500 |
| }, |
| { |
| "epoch": 10.201070690890697, |
| "grad_norm": 1.1439704895019531, |
| "learning_rate": 6.0210950563821125e-05, |
| "loss": 3.3491, |
| "step": 242000 |
| }, |
| { |
| "epoch": 10.222147283227248, |
| "grad_norm": 1.0892879962921143, |
| "learning_rate": 6.0125198302105223e-05, |
| "loss": 3.3482, |
| "step": 242500 |
| }, |
| { |
| "epoch": 10.2432238755638, |
| "grad_norm": 1.086212396621704, |
| "learning_rate": 6.0039446040389315e-05, |
| "loss": 3.345, |
| "step": 243000 |
| }, |
| { |
| "epoch": 10.26430046790035, |
| "grad_norm": 1.16683828830719, |
| "learning_rate": 5.995369377867341e-05, |
| "loss": 3.3493, |
| "step": 243500 |
| }, |
| { |
| "epoch": 10.2853770602369, |
| "grad_norm": 1.1163733005523682, |
| "learning_rate": 5.986794151695752e-05, |
| "loss": 3.3397, |
| "step": 244000 |
| }, |
| { |
| "epoch": 10.306453652573452, |
| "grad_norm": 1.2020244598388672, |
| "learning_rate": 5.97821892552416e-05, |
| "loss": 3.347, |
| "step": 244500 |
| }, |
| { |
| "epoch": 10.327530244910003, |
| "grad_norm": 1.1550841331481934, |
| "learning_rate": 5.969643699352571e-05, |
| "loss": 3.3469, |
| "step": 245000 |
| }, |
| { |
| "epoch": 10.348606837246553, |
| "grad_norm": 1.1380218267440796, |
| "learning_rate": 5.9610684731809806e-05, |
| "loss": 3.3449, |
| "step": 245500 |
| }, |
| { |
| "epoch": 10.369683429583105, |
| "grad_norm": 1.0953848361968994, |
| "learning_rate": 5.95249324700939e-05, |
| "loss": 3.3439, |
| "step": 246000 |
| }, |
| { |
| "epoch": 10.390760021919656, |
| "grad_norm": 1.1434345245361328, |
| "learning_rate": 5.9439180208377996e-05, |
| "loss": 3.3443, |
| "step": 246500 |
| }, |
| { |
| "epoch": 10.411836614256208, |
| "grad_norm": 1.1137181520462036, |
| "learning_rate": 5.93534279466621e-05, |
| "loss": 3.3412, |
| "step": 247000 |
| }, |
| { |
| "epoch": 10.432913206592758, |
| "grad_norm": 1.126103162765503, |
| "learning_rate": 5.9267675684946186e-05, |
| "loss": 3.3444, |
| "step": 247500 |
| }, |
| { |
| "epoch": 10.453989798929308, |
| "grad_norm": 1.1255971193313599, |
| "learning_rate": 5.918192342323029e-05, |
| "loss": 3.3395, |
| "step": 248000 |
| }, |
| { |
| "epoch": 10.47506639126586, |
| "grad_norm": 1.1143203973770142, |
| "learning_rate": 5.909617116151439e-05, |
| "loss": 3.3403, |
| "step": 248500 |
| }, |
| { |
| "epoch": 10.496142983602411, |
| "grad_norm": 1.1370035409927368, |
| "learning_rate": 5.901041889979848e-05, |
| "loss": 3.3376, |
| "step": 249000 |
| }, |
| { |
| "epoch": 10.517219575938963, |
| "grad_norm": 1.1377500295639038, |
| "learning_rate": 5.892466663808258e-05, |
| "loss": 3.339, |
| "step": 249500 |
| }, |
| { |
| "epoch": 10.538296168275513, |
| "grad_norm": 1.134381890296936, |
| "learning_rate": 5.8838914376366684e-05, |
| "loss": 3.341, |
| "step": 250000 |
| }, |
| { |
| "epoch": 10.538296168275513, |
| "eval_accuracy": 0.7328276784427147, |
| "eval_loss": 3.1094765663146973, |
| "eval_runtime": 205.7274, |
| "eval_samples_per_second": 586.728, |
| "eval_steps_per_second": 6.115, |
| "step": 250000 |
| }, |
| { |
| "epoch": 10.559372760612064, |
| "grad_norm": 1.118066668510437, |
| "learning_rate": 5.8753162114650776e-05, |
| "loss": 3.3435, |
| "step": 250500 |
| }, |
| { |
| "epoch": 10.580449352948616, |
| "grad_norm": 1.1390628814697266, |
| "learning_rate": 5.8667409852934874e-05, |
| "loss": 3.3378, |
| "step": 251000 |
| }, |
| { |
| "epoch": 10.601525945285166, |
| "grad_norm": 1.0764518976211548, |
| "learning_rate": 5.858165759121897e-05, |
| "loss": 3.336, |
| "step": 251500 |
| }, |
| { |
| "epoch": 10.622602537621717, |
| "grad_norm": 1.087024450302124, |
| "learning_rate": 5.8495905329503064e-05, |
| "loss": 3.3383, |
| "step": 252000 |
| }, |
| { |
| "epoch": 10.643679129958269, |
| "grad_norm": 1.1338177919387817, |
| "learning_rate": 5.841015306778716e-05, |
| "loss": 3.338, |
| "step": 252500 |
| }, |
| { |
| "epoch": 10.664755722294819, |
| "grad_norm": 1.131191372871399, |
| "learning_rate": 5.832440080607127e-05, |
| "loss": 3.3378, |
| "step": 253000 |
| }, |
| { |
| "epoch": 10.685832314631371, |
| "grad_norm": 1.1368319988250732, |
| "learning_rate": 5.823864854435536e-05, |
| "loss": 3.3322, |
| "step": 253500 |
| }, |
| { |
| "epoch": 10.706908906967922, |
| "grad_norm": 1.1224875450134277, |
| "learning_rate": 5.815289628263946e-05, |
| "loss": 3.3372, |
| "step": 254000 |
| }, |
| { |
| "epoch": 10.727985499304472, |
| "grad_norm": 1.1417977809906006, |
| "learning_rate": 5.8067144020923556e-05, |
| "loss": 3.3403, |
| "step": 254500 |
| }, |
| { |
| "epoch": 10.749062091641024, |
| "grad_norm": 1.179983139038086, |
| "learning_rate": 5.798139175920765e-05, |
| "loss": 3.3352, |
| "step": 255000 |
| }, |
| { |
| "epoch": 10.770138683977574, |
| "grad_norm": 1.1416711807250977, |
| "learning_rate": 5.789563949749175e-05, |
| "loss": 3.3373, |
| "step": 255500 |
| }, |
| { |
| "epoch": 10.791215276314126, |
| "grad_norm": 1.1273174285888672, |
| "learning_rate": 5.780988723577585e-05, |
| "loss": 3.334, |
| "step": 256000 |
| }, |
| { |
| "epoch": 10.812291868650677, |
| "grad_norm": 1.1399145126342773, |
| "learning_rate": 5.772413497405994e-05, |
| "loss": 3.3364, |
| "step": 256500 |
| }, |
| { |
| "epoch": 10.833368460987227, |
| "grad_norm": 1.0943242311477661, |
| "learning_rate": 5.763838271234404e-05, |
| "loss": 3.3319, |
| "step": 257000 |
| }, |
| { |
| "epoch": 10.85444505332378, |
| "grad_norm": 1.14568030834198, |
| "learning_rate": 5.7552630450628146e-05, |
| "loss": 3.3338, |
| "step": 257500 |
| }, |
| { |
| "epoch": 10.87552164566033, |
| "grad_norm": 1.1244406700134277, |
| "learning_rate": 5.746687818891223e-05, |
| "loss": 3.3311, |
| "step": 258000 |
| }, |
| { |
| "epoch": 10.89659823799688, |
| "grad_norm": 1.1073025465011597, |
| "learning_rate": 5.7381125927196335e-05, |
| "loss": 3.3326, |
| "step": 258500 |
| }, |
| { |
| "epoch": 10.917674830333432, |
| "grad_norm": 1.149732232093811, |
| "learning_rate": 5.7295373665480434e-05, |
| "loss": 3.3333, |
| "step": 259000 |
| }, |
| { |
| "epoch": 10.938751422669982, |
| "grad_norm": 1.1464306116104126, |
| "learning_rate": 5.7209621403764525e-05, |
| "loss": 3.3311, |
| "step": 259500 |
| }, |
| { |
| "epoch": 10.959828015006533, |
| "grad_norm": 1.1285885572433472, |
| "learning_rate": 5.7123869142048624e-05, |
| "loss": 3.3292, |
| "step": 260000 |
| }, |
| { |
| "epoch": 10.959828015006533, |
| "eval_accuracy": 0.7344874061667495, |
| "eval_loss": 3.100477457046509, |
| "eval_runtime": 208.1804, |
| "eval_samples_per_second": 579.815, |
| "eval_steps_per_second": 6.043, |
| "step": 260000 |
| }, |
| { |
| "epoch": 10.980904607343085, |
| "grad_norm": 1.1759260892868042, |
| "learning_rate": 5.703811688033273e-05, |
| "loss": 3.3254, |
| "step": 260500 |
| }, |
| { |
| "epoch": 11.001981199679635, |
| "grad_norm": 1.1321406364440918, |
| "learning_rate": 5.695236461861681e-05, |
| "loss": 3.3278, |
| "step": 261000 |
| }, |
| { |
| "epoch": 11.023057792016187, |
| "grad_norm": 1.1358826160430908, |
| "learning_rate": 5.686661235690092e-05, |
| "loss": 3.3231, |
| "step": 261500 |
| }, |
| { |
| "epoch": 11.044134384352738, |
| "grad_norm": 1.1364179849624634, |
| "learning_rate": 5.678086009518502e-05, |
| "loss": 3.3284, |
| "step": 262000 |
| }, |
| { |
| "epoch": 11.065210976689288, |
| "grad_norm": 1.163324236869812, |
| "learning_rate": 5.669510783346911e-05, |
| "loss": 3.3293, |
| "step": 262500 |
| }, |
| { |
| "epoch": 11.08628756902584, |
| "grad_norm": 1.185186743736267, |
| "learning_rate": 5.6609355571753207e-05, |
| "loss": 3.3247, |
| "step": 263000 |
| }, |
| { |
| "epoch": 11.10736416136239, |
| "grad_norm": 1.211007833480835, |
| "learning_rate": 5.652360331003731e-05, |
| "loss": 3.3262, |
| "step": 263500 |
| }, |
| { |
| "epoch": 11.128440753698943, |
| "grad_norm": 1.104915738105774, |
| "learning_rate": 5.6437851048321396e-05, |
| "loss": 3.3218, |
| "step": 264000 |
| }, |
| { |
| "epoch": 11.149517346035493, |
| "grad_norm": 1.188495397567749, |
| "learning_rate": 5.63520987866055e-05, |
| "loss": 3.3263, |
| "step": 264500 |
| }, |
| { |
| "epoch": 11.170593938372043, |
| "grad_norm": 1.099417805671692, |
| "learning_rate": 5.62663465248896e-05, |
| "loss": 3.3221, |
| "step": 265000 |
| }, |
| { |
| "epoch": 11.191670530708596, |
| "grad_norm": 1.142255187034607, |
| "learning_rate": 5.618059426317369e-05, |
| "loss": 3.3197, |
| "step": 265500 |
| }, |
| { |
| "epoch": 11.212747123045146, |
| "grad_norm": 1.13813054561615, |
| "learning_rate": 5.609484200145779e-05, |
| "loss": 3.3248, |
| "step": 266000 |
| }, |
| { |
| "epoch": 11.233823715381696, |
| "grad_norm": 1.1095728874206543, |
| "learning_rate": 5.6009089739741895e-05, |
| "loss": 3.3232, |
| "step": 266500 |
| }, |
| { |
| "epoch": 11.254900307718248, |
| "grad_norm": 1.169758677482605, |
| "learning_rate": 5.5923337478025986e-05, |
| "loss": 3.3208, |
| "step": 267000 |
| }, |
| { |
| "epoch": 11.275976900054799, |
| "grad_norm": 1.1822787523269653, |
| "learning_rate": 5.5837585216310085e-05, |
| "loss": 3.3243, |
| "step": 267500 |
| }, |
| { |
| "epoch": 11.29705349239135, |
| "grad_norm": 1.1744214296340942, |
| "learning_rate": 5.575183295459418e-05, |
| "loss": 3.3192, |
| "step": 268000 |
| }, |
| { |
| "epoch": 11.318130084727901, |
| "grad_norm": 1.122445821762085, |
| "learning_rate": 5.5666080692878274e-05, |
| "loss": 3.3227, |
| "step": 268500 |
| }, |
| { |
| "epoch": 11.339206677064452, |
| "grad_norm": 1.1298125982284546, |
| "learning_rate": 5.558032843116238e-05, |
| "loss": 3.322, |
| "step": 269000 |
| }, |
| { |
| "epoch": 11.360283269401004, |
| "grad_norm": 1.1597175598144531, |
| "learning_rate": 5.549457616944648e-05, |
| "loss": 3.321, |
| "step": 269500 |
| }, |
| { |
| "epoch": 11.381359861737554, |
| "grad_norm": 1.1467528343200684, |
| "learning_rate": 5.540882390773057e-05, |
| "loss": 3.3173, |
| "step": 270000 |
| }, |
| { |
| "epoch": 11.381359861737554, |
| "eval_accuracy": 0.7354613491453346, |
| "eval_loss": 3.094395875930786, |
| "eval_runtime": 216.6229, |
| "eval_samples_per_second": 557.217, |
| "eval_steps_per_second": 5.807, |
| "step": 270000 |
| }, |
| { |
| "epoch": 11.402436454074106, |
| "grad_norm": 1.1795012950897217, |
| "learning_rate": 5.532307164601467e-05, |
| "loss": 3.3199, |
| "step": 270500 |
| }, |
| { |
| "epoch": 11.423513046410656, |
| "grad_norm": 1.1417827606201172, |
| "learning_rate": 5.5237319384298766e-05, |
| "loss": 3.3184, |
| "step": 271000 |
| }, |
| { |
| "epoch": 11.444589638747207, |
| "grad_norm": 1.1047520637512207, |
| "learning_rate": 5.515156712258286e-05, |
| "loss": 3.3153, |
| "step": 271500 |
| }, |
| { |
| "epoch": 11.465666231083759, |
| "grad_norm": 1.1155893802642822, |
| "learning_rate": 5.506581486086696e-05, |
| "loss": 3.3167, |
| "step": 272000 |
| }, |
| { |
| "epoch": 11.48674282342031, |
| "grad_norm": 1.112353801727295, |
| "learning_rate": 5.498006259915105e-05, |
| "loss": 3.3179, |
| "step": 272500 |
| }, |
| { |
| "epoch": 11.50781941575686, |
| "grad_norm": 1.1469171047210693, |
| "learning_rate": 5.489431033743515e-05, |
| "loss": 3.3159, |
| "step": 273000 |
| }, |
| { |
| "epoch": 11.528896008093412, |
| "grad_norm": 1.1078561544418335, |
| "learning_rate": 5.480855807571925e-05, |
| "loss": 3.3202, |
| "step": 273500 |
| }, |
| { |
| "epoch": 11.549972600429962, |
| "grad_norm": 1.1520354747772217, |
| "learning_rate": 5.472280581400334e-05, |
| "loss": 3.3152, |
| "step": 274000 |
| }, |
| { |
| "epoch": 11.571049192766514, |
| "grad_norm": 1.1484616994857788, |
| "learning_rate": 5.463705355228744e-05, |
| "loss": 3.315, |
| "step": 274500 |
| }, |
| { |
| "epoch": 11.592125785103065, |
| "grad_norm": 1.11712646484375, |
| "learning_rate": 5.4551301290571546e-05, |
| "loss": 3.3179, |
| "step": 275000 |
| }, |
| { |
| "epoch": 11.613202377439615, |
| "grad_norm": 1.1272873878479004, |
| "learning_rate": 5.446554902885563e-05, |
| "loss": 3.3154, |
| "step": 275500 |
| }, |
| { |
| "epoch": 11.634278969776167, |
| "grad_norm": 1.104549527168274, |
| "learning_rate": 5.4379796767139736e-05, |
| "loss": 3.3144, |
| "step": 276000 |
| }, |
| { |
| "epoch": 11.655355562112717, |
| "grad_norm": 1.153380274772644, |
| "learning_rate": 5.4294044505423834e-05, |
| "loss": 3.3141, |
| "step": 276500 |
| }, |
| { |
| "epoch": 11.67643215444927, |
| "grad_norm": 1.1044162511825562, |
| "learning_rate": 5.4208292243707925e-05, |
| "loss": 3.3133, |
| "step": 277000 |
| }, |
| { |
| "epoch": 11.69750874678582, |
| "grad_norm": 1.1153321266174316, |
| "learning_rate": 5.4122539981992024e-05, |
| "loss": 3.3143, |
| "step": 277500 |
| }, |
| { |
| "epoch": 11.71858533912237, |
| "grad_norm": 1.0939761400222778, |
| "learning_rate": 5.403678772027613e-05, |
| "loss": 3.3089, |
| "step": 278000 |
| }, |
| { |
| "epoch": 11.739661931458922, |
| "grad_norm": 1.1236324310302734, |
| "learning_rate": 5.395103545856022e-05, |
| "loss": 3.3118, |
| "step": 278500 |
| }, |
| { |
| "epoch": 11.760738523795473, |
| "grad_norm": 1.1610225439071655, |
| "learning_rate": 5.386528319684432e-05, |
| "loss": 3.3108, |
| "step": 279000 |
| }, |
| { |
| "epoch": 11.781815116132023, |
| "grad_norm": 1.1525251865386963, |
| "learning_rate": 5.377953093512842e-05, |
| "loss": 3.3067, |
| "step": 279500 |
| }, |
| { |
| "epoch": 11.802891708468575, |
| "grad_norm": 1.1421979665756226, |
| "learning_rate": 5.369377867341251e-05, |
| "loss": 3.3162, |
| "step": 280000 |
| }, |
| { |
| "epoch": 11.802891708468575, |
| "eval_accuracy": 0.737312036395569, |
| "eval_loss": 3.083812713623047, |
| "eval_runtime": 206.7473, |
| "eval_samples_per_second": 583.833, |
| "eval_steps_per_second": 6.085, |
| "step": 280000 |
| }, |
| { |
| "epoch": 11.823968300805126, |
| "grad_norm": 1.1560571193695068, |
| "learning_rate": 5.3608026411696614e-05, |
| "loss": 3.3157, |
| "step": 280500 |
| }, |
| { |
| "epoch": 11.845044893141678, |
| "grad_norm": 1.1876248121261597, |
| "learning_rate": 5.352227414998071e-05, |
| "loss": 3.3055, |
| "step": 281000 |
| }, |
| { |
| "epoch": 11.866121485478228, |
| "grad_norm": 1.164538025856018, |
| "learning_rate": 5.34365218882648e-05, |
| "loss": 3.309, |
| "step": 281500 |
| }, |
| { |
| "epoch": 11.887198077814778, |
| "grad_norm": 1.1904025077819824, |
| "learning_rate": 5.33507696265489e-05, |
| "loss": 3.3038, |
| "step": 282000 |
| }, |
| { |
| "epoch": 11.90827467015133, |
| "grad_norm": 1.1061123609542847, |
| "learning_rate": 5.3265017364833e-05, |
| "loss": 3.3125, |
| "step": 282500 |
| }, |
| { |
| "epoch": 11.92935126248788, |
| "grad_norm": 1.177546739578247, |
| "learning_rate": 5.317926510311709e-05, |
| "loss": 3.3071, |
| "step": 283000 |
| }, |
| { |
| "epoch": 11.950427854824433, |
| "grad_norm": 1.1350585222244263, |
| "learning_rate": 5.3093512841401197e-05, |
| "loss": 3.307, |
| "step": 283500 |
| }, |
| { |
| "epoch": 11.971504447160983, |
| "grad_norm": 1.168764591217041, |
| "learning_rate": 5.3007760579685295e-05, |
| "loss": 3.3093, |
| "step": 284000 |
| }, |
| { |
| "epoch": 11.992581039497534, |
| "grad_norm": 1.1399846076965332, |
| "learning_rate": 5.2922008317969386e-05, |
| "loss": 3.3124, |
| "step": 284500 |
| }, |
| { |
| "epoch": 12.013657631834086, |
| "grad_norm": 1.1211718320846558, |
| "learning_rate": 5.2836256056253485e-05, |
| "loss": 3.3064, |
| "step": 285000 |
| }, |
| { |
| "epoch": 12.034734224170636, |
| "grad_norm": 1.1564571857452393, |
| "learning_rate": 5.275050379453759e-05, |
| "loss": 3.3074, |
| "step": 285500 |
| }, |
| { |
| "epoch": 12.055810816507186, |
| "grad_norm": 1.0877307653427124, |
| "learning_rate": 5.2664751532821675e-05, |
| "loss": 3.3035, |
| "step": 286000 |
| }, |
| { |
| "epoch": 12.076887408843739, |
| "grad_norm": 1.142830491065979, |
| "learning_rate": 5.257899927110578e-05, |
| "loss": 3.3035, |
| "step": 286500 |
| }, |
| { |
| "epoch": 12.097964001180289, |
| "grad_norm": 1.1289803981781006, |
| "learning_rate": 5.249324700938988e-05, |
| "loss": 3.3008, |
| "step": 287000 |
| }, |
| { |
| "epoch": 12.119040593516841, |
| "grad_norm": 1.1464356184005737, |
| "learning_rate": 5.240749474767397e-05, |
| "loss": 3.2987, |
| "step": 287500 |
| }, |
| { |
| "epoch": 12.140117185853391, |
| "grad_norm": 1.1905564069747925, |
| "learning_rate": 5.232174248595807e-05, |
| "loss": 3.3028, |
| "step": 288000 |
| }, |
| { |
| "epoch": 12.161193778189942, |
| "grad_norm": 1.1306829452514648, |
| "learning_rate": 5.223599022424217e-05, |
| "loss": 3.3037, |
| "step": 288500 |
| }, |
| { |
| "epoch": 12.182270370526494, |
| "grad_norm": 1.1403478384017944, |
| "learning_rate": 5.215023796252626e-05, |
| "loss": 3.3072, |
| "step": 289000 |
| }, |
| { |
| "epoch": 12.203346962863044, |
| "grad_norm": 1.1387965679168701, |
| "learning_rate": 5.206448570081036e-05, |
| "loss": 3.299, |
| "step": 289500 |
| }, |
| { |
| "epoch": 12.224423555199595, |
| "grad_norm": 1.1716845035552979, |
| "learning_rate": 5.197873343909446e-05, |
| "loss": 3.298, |
| "step": 290000 |
| }, |
| { |
| "epoch": 12.224423555199595, |
| "eval_accuracy": 0.738551636960977, |
| "eval_loss": 3.076448917388916, |
| "eval_runtime": 202.5634, |
| "eval_samples_per_second": 595.893, |
| "eval_steps_per_second": 6.21, |
| "step": 290000 |
| }, |
| { |
| "epoch": 12.245500147536147, |
| "grad_norm": 1.1910455226898193, |
| "learning_rate": 5.189298117737855e-05, |
| "loss": 3.2996, |
| "step": 290500 |
| }, |
| { |
| "epoch": 12.266576739872697, |
| "grad_norm": 1.151992678642273, |
| "learning_rate": 5.180722891566265e-05, |
| "loss": 3.2976, |
| "step": 291000 |
| }, |
| { |
| "epoch": 12.28765333220925, |
| "grad_norm": 1.1691997051239014, |
| "learning_rate": 5.1721476653946756e-05, |
| "loss": 3.2969, |
| "step": 291500 |
| }, |
| { |
| "epoch": 12.3087299245458, |
| "grad_norm": 1.1572399139404297, |
| "learning_rate": 5.163572439223085e-05, |
| "loss": 3.3006, |
| "step": 292000 |
| }, |
| { |
| "epoch": 12.32980651688235, |
| "grad_norm": 1.1561634540557861, |
| "learning_rate": 5.1549972130514946e-05, |
| "loss": 3.3, |
| "step": 292500 |
| }, |
| { |
| "epoch": 12.350883109218902, |
| "grad_norm": 1.169138789176941, |
| "learning_rate": 5.1464219868799044e-05, |
| "loss": 3.2986, |
| "step": 293000 |
| }, |
| { |
| "epoch": 12.371959701555452, |
| "grad_norm": 1.1753603219985962, |
| "learning_rate": 5.1378467607083136e-05, |
| "loss": 3.2966, |
| "step": 293500 |
| }, |
| { |
| "epoch": 12.393036293892003, |
| "grad_norm": 1.1309871673583984, |
| "learning_rate": 5.1292715345367234e-05, |
| "loss": 3.2977, |
| "step": 294000 |
| }, |
| { |
| "epoch": 12.414112886228555, |
| "grad_norm": 1.1298264265060425, |
| "learning_rate": 5.120696308365134e-05, |
| "loss": 3.2947, |
| "step": 294500 |
| }, |
| { |
| "epoch": 12.435189478565105, |
| "grad_norm": 1.1679694652557373, |
| "learning_rate": 5.112121082193543e-05, |
| "loss": 3.2996, |
| "step": 295000 |
| }, |
| { |
| "epoch": 12.456266070901657, |
| "grad_norm": 1.1328699588775635, |
| "learning_rate": 5.103545856021953e-05, |
| "loss": 3.2968, |
| "step": 295500 |
| }, |
| { |
| "epoch": 12.477342663238208, |
| "grad_norm": 1.1515026092529297, |
| "learning_rate": 5.094970629850363e-05, |
| "loss": 3.2958, |
| "step": 296000 |
| }, |
| { |
| "epoch": 12.498419255574758, |
| "grad_norm": 1.1479114294052124, |
| "learning_rate": 5.086395403678772e-05, |
| "loss": 3.2961, |
| "step": 296500 |
| }, |
| { |
| "epoch": 12.51949584791131, |
| "grad_norm": 1.1606543064117432, |
| "learning_rate": 5.0778201775071824e-05, |
| "loss": 3.3014, |
| "step": 297000 |
| }, |
| { |
| "epoch": 12.54057244024786, |
| "grad_norm": 1.1763315200805664, |
| "learning_rate": 5.069244951335592e-05, |
| "loss": 3.2985, |
| "step": 297500 |
| }, |
| { |
| "epoch": 12.561649032584413, |
| "grad_norm": 1.1756614446640015, |
| "learning_rate": 5.0606697251640014e-05, |
| "loss": 3.2925, |
| "step": 298000 |
| }, |
| { |
| "epoch": 12.582725624920963, |
| "grad_norm": 1.1701266765594482, |
| "learning_rate": 5.052094498992411e-05, |
| "loss": 3.2965, |
| "step": 298500 |
| }, |
| { |
| "epoch": 12.603802217257513, |
| "grad_norm": 1.1340301036834717, |
| "learning_rate": 5.043519272820822e-05, |
| "loss": 3.2981, |
| "step": 299000 |
| }, |
| { |
| "epoch": 12.624878809594065, |
| "grad_norm": 1.1425764560699463, |
| "learning_rate": 5.03494404664923e-05, |
| "loss": 3.2915, |
| "step": 299500 |
| }, |
| { |
| "epoch": 12.645955401930616, |
| "grad_norm": 1.165718674659729, |
| "learning_rate": 5.026368820477641e-05, |
| "loss": 3.293, |
| "step": 300000 |
| }, |
| { |
| "epoch": 12.645955401930616, |
| "eval_accuracy": 0.7399510983559554, |
| "eval_loss": 3.0675549507141113, |
| "eval_runtime": 205.3957, |
| "eval_samples_per_second": 587.676, |
| "eval_steps_per_second": 6.125, |
| "step": 300000 |
| }, |
| { |
| "epoch": 12.667031994267166, |
| "grad_norm": 1.1546664237976074, |
| "learning_rate": 5.0177935943060505e-05, |
| "loss": 3.2886, |
| "step": 300500 |
| }, |
| { |
| "epoch": 12.688108586603718, |
| "grad_norm": 1.1368634700775146, |
| "learning_rate": 5.00921836813446e-05, |
| "loss": 3.2965, |
| "step": 301000 |
| }, |
| { |
| "epoch": 12.709185178940269, |
| "grad_norm": 1.1606389284133911, |
| "learning_rate": 5.0006431419628695e-05, |
| "loss": 3.2939, |
| "step": 301500 |
| }, |
| { |
| "epoch": 12.73026177127682, |
| "grad_norm": 1.1257100105285645, |
| "learning_rate": 4.992067915791279e-05, |
| "loss": 3.2932, |
| "step": 302000 |
| }, |
| { |
| "epoch": 12.751338363613371, |
| "grad_norm": 1.1844743490219116, |
| "learning_rate": 4.9834926896196885e-05, |
| "loss": 3.2878, |
| "step": 302500 |
| }, |
| { |
| "epoch": 12.772414955949921, |
| "grad_norm": 1.178240418434143, |
| "learning_rate": 4.974917463448099e-05, |
| "loss": 3.2952, |
| "step": 303000 |
| }, |
| { |
| "epoch": 12.793491548286474, |
| "grad_norm": 1.127692461013794, |
| "learning_rate": 4.966342237276508e-05, |
| "loss": 3.2944, |
| "step": 303500 |
| }, |
| { |
| "epoch": 12.814568140623024, |
| "grad_norm": 1.1594480276107788, |
| "learning_rate": 4.957767011104918e-05, |
| "loss": 3.2931, |
| "step": 304000 |
| }, |
| { |
| "epoch": 12.835644732959576, |
| "grad_norm": 1.1510993242263794, |
| "learning_rate": 4.949191784933328e-05, |
| "loss": 3.2905, |
| "step": 304500 |
| }, |
| { |
| "epoch": 12.856721325296126, |
| "grad_norm": 1.1831223964691162, |
| "learning_rate": 4.9406165587617376e-05, |
| "loss": 3.2908, |
| "step": 305000 |
| }, |
| { |
| "epoch": 12.877797917632677, |
| "grad_norm": 1.1877171993255615, |
| "learning_rate": 4.932041332590147e-05, |
| "loss": 3.2938, |
| "step": 305500 |
| }, |
| { |
| "epoch": 12.898874509969229, |
| "grad_norm": 1.2035733461380005, |
| "learning_rate": 4.923466106418557e-05, |
| "loss": 3.2887, |
| "step": 306000 |
| }, |
| { |
| "epoch": 12.91995110230578, |
| "grad_norm": 1.1415528059005737, |
| "learning_rate": 4.9148908802469665e-05, |
| "loss": 3.2873, |
| "step": 306500 |
| }, |
| { |
| "epoch": 12.94102769464233, |
| "grad_norm": 1.1466418504714966, |
| "learning_rate": 4.906315654075376e-05, |
| "loss": 3.2895, |
| "step": 307000 |
| }, |
| { |
| "epoch": 12.962104286978882, |
| "grad_norm": 1.1475776433944702, |
| "learning_rate": 4.897740427903786e-05, |
| "loss": 3.2932, |
| "step": 307500 |
| }, |
| { |
| "epoch": 12.983180879315432, |
| "grad_norm": 1.232168436050415, |
| "learning_rate": 4.889165201732196e-05, |
| "loss": 3.2934, |
| "step": 308000 |
| }, |
| { |
| "epoch": 13.004257471651984, |
| "grad_norm": 1.1670901775360107, |
| "learning_rate": 4.880589975560606e-05, |
| "loss": 3.2876, |
| "step": 308500 |
| }, |
| { |
| "epoch": 13.025334063988534, |
| "grad_norm": 1.147477626800537, |
| "learning_rate": 4.8720147493890156e-05, |
| "loss": 3.2809, |
| "step": 309000 |
| }, |
| { |
| "epoch": 13.046410656325085, |
| "grad_norm": 1.144195318222046, |
| "learning_rate": 4.8634395232174254e-05, |
| "loss": 3.2859, |
| "step": 309500 |
| }, |
| { |
| "epoch": 13.067487248661637, |
| "grad_norm": 1.127899408340454, |
| "learning_rate": 4.8548642970458346e-05, |
| "loss": 3.2808, |
| "step": 310000 |
| }, |
| { |
| "epoch": 13.067487248661637, |
| "eval_accuracy": 0.7413039897467341, |
| "eval_loss": 3.0613441467285156, |
| "eval_runtime": 201.9806, |
| "eval_samples_per_second": 597.612, |
| "eval_steps_per_second": 6.228, |
| "step": 310000 |
| }, |
| { |
| "epoch": 13.088563840998187, |
| "grad_norm": 1.125076413154602, |
| "learning_rate": 4.846289070874245e-05, |
| "loss": 3.2847, |
| "step": 310500 |
| }, |
| { |
| "epoch": 13.109640433334738, |
| "grad_norm": 1.1331325769424438, |
| "learning_rate": 4.837713844702654e-05, |
| "loss": 3.2853, |
| "step": 311000 |
| }, |
| { |
| "epoch": 13.13071702567129, |
| "grad_norm": 1.1389740705490112, |
| "learning_rate": 4.829138618531064e-05, |
| "loss": 3.2812, |
| "step": 311500 |
| }, |
| { |
| "epoch": 13.15179361800784, |
| "grad_norm": 1.1709684133529663, |
| "learning_rate": 4.820563392359474e-05, |
| "loss": 3.2826, |
| "step": 312000 |
| }, |
| { |
| "epoch": 13.172870210344392, |
| "grad_norm": 1.1737103462219238, |
| "learning_rate": 4.811988166187884e-05, |
| "loss": 3.28, |
| "step": 312500 |
| }, |
| { |
| "epoch": 13.193946802680943, |
| "grad_norm": 1.1896841526031494, |
| "learning_rate": 4.803412940016293e-05, |
| "loss": 3.2814, |
| "step": 313000 |
| }, |
| { |
| "epoch": 13.215023395017493, |
| "grad_norm": 1.1704668998718262, |
| "learning_rate": 4.7948377138447034e-05, |
| "loss": 3.2837, |
| "step": 313500 |
| }, |
| { |
| "epoch": 13.236099987354045, |
| "grad_norm": 1.1782784461975098, |
| "learning_rate": 4.7862624876731126e-05, |
| "loss": 3.2782, |
| "step": 314000 |
| }, |
| { |
| "epoch": 13.257176579690595, |
| "grad_norm": 1.1247632503509521, |
| "learning_rate": 4.7776872615015224e-05, |
| "loss": 3.2806, |
| "step": 314500 |
| }, |
| { |
| "epoch": 13.278253172027147, |
| "grad_norm": 1.148779273033142, |
| "learning_rate": 4.7691120353299315e-05, |
| "loss": 3.2772, |
| "step": 315000 |
| }, |
| { |
| "epoch": 13.299329764363698, |
| "grad_norm": 1.1245955228805542, |
| "learning_rate": 4.760536809158342e-05, |
| "loss": 3.2846, |
| "step": 315500 |
| }, |
| { |
| "epoch": 13.320406356700248, |
| "grad_norm": 1.1448092460632324, |
| "learning_rate": 4.751961582986751e-05, |
| "loss": 3.2813, |
| "step": 316000 |
| }, |
| { |
| "epoch": 13.3414829490368, |
| "grad_norm": 1.1150004863739014, |
| "learning_rate": 4.743386356815161e-05, |
| "loss": 3.2832, |
| "step": 316500 |
| }, |
| { |
| "epoch": 13.36255954137335, |
| "grad_norm": 1.1444592475891113, |
| "learning_rate": 4.734811130643571e-05, |
| "loss": 3.2798, |
| "step": 317000 |
| }, |
| { |
| "epoch": 13.383636133709901, |
| "grad_norm": 1.1749241352081299, |
| "learning_rate": 4.726235904471981e-05, |
| "loss": 3.2771, |
| "step": 317500 |
| }, |
| { |
| "epoch": 13.404712726046453, |
| "grad_norm": 1.2211580276489258, |
| "learning_rate": 4.71766067830039e-05, |
| "loss": 3.2795, |
| "step": 318000 |
| }, |
| { |
| "epoch": 13.425789318383003, |
| "grad_norm": 1.1678978204727173, |
| "learning_rate": 4.7090854521288004e-05, |
| "loss": 3.2828, |
| "step": 318500 |
| }, |
| { |
| "epoch": 13.446865910719556, |
| "grad_norm": 1.1619858741760254, |
| "learning_rate": 4.7005102259572095e-05, |
| "loss": 3.278, |
| "step": 319000 |
| }, |
| { |
| "epoch": 13.467942503056106, |
| "grad_norm": 1.2026293277740479, |
| "learning_rate": 4.6919349997856193e-05, |
| "loss": 3.2796, |
| "step": 319500 |
| }, |
| { |
| "epoch": 13.489019095392656, |
| "grad_norm": 1.1696882247924805, |
| "learning_rate": 4.683359773614029e-05, |
| "loss": 3.2786, |
| "step": 320000 |
| }, |
| { |
| "epoch": 13.489019095392656, |
| "eval_accuracy": 0.7422862191210289, |
| "eval_loss": 3.0541787147521973, |
| "eval_runtime": 202.9471, |
| "eval_samples_per_second": 594.766, |
| "eval_steps_per_second": 6.199, |
| "step": 320000 |
| }, |
| { |
| "epoch": 13.510095687729208, |
| "grad_norm": 1.167097806930542, |
| "learning_rate": 4.674784547442439e-05, |
| "loss": 3.2826, |
| "step": 320500 |
| }, |
| { |
| "epoch": 13.531172280065759, |
| "grad_norm": 1.1454930305480957, |
| "learning_rate": 4.666209321270849e-05, |
| "loss": 3.2784, |
| "step": 321000 |
| }, |
| { |
| "epoch": 13.55224887240231, |
| "grad_norm": 1.1719410419464111, |
| "learning_rate": 4.657634095099259e-05, |
| "loss": 3.2791, |
| "step": 321500 |
| }, |
| { |
| "epoch": 13.573325464738861, |
| "grad_norm": 1.1700538396835327, |
| "learning_rate": 4.6490588689276685e-05, |
| "loss": 3.2785, |
| "step": 322000 |
| }, |
| { |
| "epoch": 13.594402057075412, |
| "grad_norm": 1.1331679821014404, |
| "learning_rate": 4.6404836427560777e-05, |
| "loss": 3.2794, |
| "step": 322500 |
| }, |
| { |
| "epoch": 13.615478649411964, |
| "grad_norm": 1.2087810039520264, |
| "learning_rate": 4.6319084165844875e-05, |
| "loss": 3.2723, |
| "step": 323000 |
| }, |
| { |
| "epoch": 13.636555241748514, |
| "grad_norm": 1.1538053750991821, |
| "learning_rate": 4.623333190412897e-05, |
| "loss": 3.2783, |
| "step": 323500 |
| }, |
| { |
| "epoch": 13.657631834085064, |
| "grad_norm": 1.1657546758651733, |
| "learning_rate": 4.614757964241307e-05, |
| "loss": 3.2743, |
| "step": 324000 |
| }, |
| { |
| "epoch": 13.678708426421617, |
| "grad_norm": 1.1276378631591797, |
| "learning_rate": 4.606182738069717e-05, |
| "loss": 3.2757, |
| "step": 324500 |
| }, |
| { |
| "epoch": 13.699785018758167, |
| "grad_norm": 1.1900725364685059, |
| "learning_rate": 4.597607511898127e-05, |
| "loss": 3.2762, |
| "step": 325000 |
| }, |
| { |
| "epoch": 13.720861611094719, |
| "grad_norm": 1.1592153310775757, |
| "learning_rate": 4.589032285726536e-05, |
| "loss": 3.2743, |
| "step": 325500 |
| }, |
| { |
| "epoch": 13.74193820343127, |
| "grad_norm": 1.1564064025878906, |
| "learning_rate": 4.5804570595549465e-05, |
| "loss": 3.2749, |
| "step": 326000 |
| }, |
| { |
| "epoch": 13.76301479576782, |
| "grad_norm": 1.185072660446167, |
| "learning_rate": 4.5718818333833556e-05, |
| "loss": 3.2785, |
| "step": 326500 |
| }, |
| { |
| "epoch": 13.784091388104372, |
| "grad_norm": 1.1844220161437988, |
| "learning_rate": 4.5633066072117655e-05, |
| "loss": 3.2768, |
| "step": 327000 |
| }, |
| { |
| "epoch": 13.805167980440922, |
| "grad_norm": 1.1187955141067505, |
| "learning_rate": 4.554731381040175e-05, |
| "loss": 3.2711, |
| "step": 327500 |
| }, |
| { |
| "epoch": 13.826244572777473, |
| "grad_norm": 1.175713062286377, |
| "learning_rate": 4.546156154868585e-05, |
| "loss": 3.2745, |
| "step": 328000 |
| }, |
| { |
| "epoch": 13.847321165114025, |
| "grad_norm": 1.1979278326034546, |
| "learning_rate": 4.537580928696994e-05, |
| "loss": 3.2744, |
| "step": 328500 |
| }, |
| { |
| "epoch": 13.868397757450575, |
| "grad_norm": 1.15327787399292, |
| "learning_rate": 4.529005702525405e-05, |
| "loss": 3.2723, |
| "step": 329000 |
| }, |
| { |
| "epoch": 13.889474349787127, |
| "grad_norm": 1.173656702041626, |
| "learning_rate": 4.520430476353814e-05, |
| "loss": 3.2709, |
| "step": 329500 |
| }, |
| { |
| "epoch": 13.910550942123677, |
| "grad_norm": 1.190041422843933, |
| "learning_rate": 4.511855250182224e-05, |
| "loss": 3.2717, |
| "step": 330000 |
| }, |
| { |
| "epoch": 13.910550942123677, |
| "eval_accuracy": 0.7431622607327596, |
| "eval_loss": 3.050483465194702, |
| "eval_runtime": 206.8733, |
| "eval_samples_per_second": 583.478, |
| "eval_steps_per_second": 6.081, |
| "step": 330000 |
| }, |
| { |
| "epoch": 13.931627534460228, |
| "grad_norm": 1.1565124988555908, |
| "learning_rate": 4.5032800240106336e-05, |
| "loss": 3.2731, |
| "step": 330500 |
| }, |
| { |
| "epoch": 13.95270412679678, |
| "grad_norm": 1.1900709867477417, |
| "learning_rate": 4.4947047978390434e-05, |
| "loss": 3.2752, |
| "step": 331000 |
| }, |
| { |
| "epoch": 13.97378071913333, |
| "grad_norm": 1.1843534708023071, |
| "learning_rate": 4.4861295716674526e-05, |
| "loss": 3.2735, |
| "step": 331500 |
| }, |
| { |
| "epoch": 13.994857311469882, |
| "grad_norm": 1.206402063369751, |
| "learning_rate": 4.4775543454958624e-05, |
| "loss": 3.2739, |
| "step": 332000 |
| }, |
| { |
| "epoch": 14.015933903806433, |
| "grad_norm": 1.192505121231079, |
| "learning_rate": 4.468979119324272e-05, |
| "loss": 3.2667, |
| "step": 332500 |
| }, |
| { |
| "epoch": 14.037010496142983, |
| "grad_norm": 1.1465531587600708, |
| "learning_rate": 4.460403893152682e-05, |
| "loss": 3.2664, |
| "step": 333000 |
| }, |
| { |
| "epoch": 14.058087088479535, |
| "grad_norm": 1.1538958549499512, |
| "learning_rate": 4.451828666981092e-05, |
| "loss": 3.2668, |
| "step": 333500 |
| }, |
| { |
| "epoch": 14.079163680816086, |
| "grad_norm": 1.1610597372055054, |
| "learning_rate": 4.443253440809502e-05, |
| "loss": 3.2669, |
| "step": 334000 |
| }, |
| { |
| "epoch": 14.100240273152636, |
| "grad_norm": 1.1734635829925537, |
| "learning_rate": 4.4346782146379116e-05, |
| "loss": 3.2661, |
| "step": 334500 |
| }, |
| { |
| "epoch": 14.121316865489188, |
| "grad_norm": 1.159870982170105, |
| "learning_rate": 4.426102988466321e-05, |
| "loss": 3.2638, |
| "step": 335000 |
| }, |
| { |
| "epoch": 14.142393457825738, |
| "grad_norm": 1.1902533769607544, |
| "learning_rate": 4.4175277622947305e-05, |
| "loss": 3.2598, |
| "step": 335500 |
| }, |
| { |
| "epoch": 14.16347005016229, |
| "grad_norm": 1.1931352615356445, |
| "learning_rate": 4.4089525361231404e-05, |
| "loss": 3.2646, |
| "step": 336000 |
| }, |
| { |
| "epoch": 14.184546642498841, |
| "grad_norm": 1.143208622932434, |
| "learning_rate": 4.40037730995155e-05, |
| "loss": 3.2602, |
| "step": 336500 |
| }, |
| { |
| "epoch": 14.205623234835391, |
| "grad_norm": 1.1963354349136353, |
| "learning_rate": 4.39180208377996e-05, |
| "loss": 3.2651, |
| "step": 337000 |
| }, |
| { |
| "epoch": 14.226699827171943, |
| "grad_norm": 1.1712560653686523, |
| "learning_rate": 4.38322685760837e-05, |
| "loss": 3.2654, |
| "step": 337500 |
| }, |
| { |
| "epoch": 14.247776419508494, |
| "grad_norm": 1.1791725158691406, |
| "learning_rate": 4.374651631436779e-05, |
| "loss": 3.2655, |
| "step": 338000 |
| }, |
| { |
| "epoch": 14.268853011845044, |
| "grad_norm": 1.1490861177444458, |
| "learning_rate": 4.3660764052651895e-05, |
| "loss": 3.2664, |
| "step": 338500 |
| }, |
| { |
| "epoch": 14.289929604181596, |
| "grad_norm": 1.2244560718536377, |
| "learning_rate": 4.357501179093599e-05, |
| "loss": 3.265, |
| "step": 339000 |
| }, |
| { |
| "epoch": 14.311006196518147, |
| "grad_norm": 1.2384368181228638, |
| "learning_rate": 4.3489259529220085e-05, |
| "loss": 3.2639, |
| "step": 339500 |
| }, |
| { |
| "epoch": 14.332082788854699, |
| "grad_norm": 1.2463946342468262, |
| "learning_rate": 4.3403507267504183e-05, |
| "loss": 3.266, |
| "step": 340000 |
| }, |
| { |
| "epoch": 14.332082788854699, |
| "eval_accuracy": 0.7444987703291956, |
| "eval_loss": 3.0426642894744873, |
| "eval_runtime": 188.6811, |
| "eval_samples_per_second": 639.735, |
| "eval_steps_per_second": 6.667, |
| "step": 340000 |
| }, |
| { |
| "epoch": 14.353159381191249, |
| "grad_norm": 1.1417078971862793, |
| "learning_rate": 4.331775500578828e-05, |
| "loss": 3.2625, |
| "step": 340500 |
| }, |
| { |
| "epoch": 14.3742359735278, |
| "grad_norm": 1.160050392150879, |
| "learning_rate": 4.323200274407237e-05, |
| "loss": 3.262, |
| "step": 341000 |
| }, |
| { |
| "epoch": 14.395312565864351, |
| "grad_norm": 1.218281626701355, |
| "learning_rate": 4.314625048235648e-05, |
| "loss": 3.2606, |
| "step": 341500 |
| }, |
| { |
| "epoch": 14.416389158200902, |
| "grad_norm": 1.2003798484802246, |
| "learning_rate": 4.306049822064057e-05, |
| "loss": 3.2647, |
| "step": 342000 |
| }, |
| { |
| "epoch": 14.437465750537454, |
| "grad_norm": 1.1490397453308105, |
| "learning_rate": 4.297474595892467e-05, |
| "loss": 3.2635, |
| "step": 342500 |
| }, |
| { |
| "epoch": 14.458542342874004, |
| "grad_norm": 1.1432554721832275, |
| "learning_rate": 4.2888993697208767e-05, |
| "loss": 3.2596, |
| "step": 343000 |
| }, |
| { |
| "epoch": 14.479618935210555, |
| "grad_norm": 1.1216256618499756, |
| "learning_rate": 4.2803241435492865e-05, |
| "loss": 3.2624, |
| "step": 343500 |
| }, |
| { |
| "epoch": 14.500695527547107, |
| "grad_norm": 1.1562144756317139, |
| "learning_rate": 4.2717489173776956e-05, |
| "loss": 3.2619, |
| "step": 344000 |
| }, |
| { |
| "epoch": 14.521772119883657, |
| "grad_norm": 1.1353654861450195, |
| "learning_rate": 4.263173691206106e-05, |
| "loss": 3.2599, |
| "step": 344500 |
| }, |
| { |
| "epoch": 14.542848712220207, |
| "grad_norm": 1.1956453323364258, |
| "learning_rate": 4.254598465034515e-05, |
| "loss": 3.2629, |
| "step": 345000 |
| }, |
| { |
| "epoch": 14.56392530455676, |
| "grad_norm": 1.1945847272872925, |
| "learning_rate": 4.246023238862925e-05, |
| "loss": 3.2593, |
| "step": 345500 |
| }, |
| { |
| "epoch": 14.58500189689331, |
| "grad_norm": 1.1688573360443115, |
| "learning_rate": 4.237448012691335e-05, |
| "loss": 3.2606, |
| "step": 346000 |
| }, |
| { |
| "epoch": 14.606078489229862, |
| "grad_norm": 1.229426622390747, |
| "learning_rate": 4.228872786519745e-05, |
| "loss": 3.2586, |
| "step": 346500 |
| }, |
| { |
| "epoch": 14.627155081566412, |
| "grad_norm": 1.1840457916259766, |
| "learning_rate": 4.220297560348154e-05, |
| "loss": 3.2615, |
| "step": 347000 |
| }, |
| { |
| "epoch": 14.648231673902963, |
| "grad_norm": 1.1648441553115845, |
| "learning_rate": 4.2117223341765645e-05, |
| "loss": 3.2595, |
| "step": 347500 |
| }, |
| { |
| "epoch": 14.669308266239515, |
| "grad_norm": 1.1526823043823242, |
| "learning_rate": 4.2031471080049736e-05, |
| "loss": 3.2593, |
| "step": 348000 |
| }, |
| { |
| "epoch": 14.690384858576065, |
| "grad_norm": 1.1709331274032593, |
| "learning_rate": 4.1945718818333834e-05, |
| "loss": 3.2594, |
| "step": 348500 |
| }, |
| { |
| "epoch": 14.711461450912616, |
| "grad_norm": 1.5731021165847778, |
| "learning_rate": 4.185996655661793e-05, |
| "loss": 3.2609, |
| "step": 349000 |
| }, |
| { |
| "epoch": 14.732538043249168, |
| "grad_norm": 1.2158201932907104, |
| "learning_rate": 4.177421429490203e-05, |
| "loss": 3.2623, |
| "step": 349500 |
| }, |
| { |
| "epoch": 14.753614635585718, |
| "grad_norm": 1.205597996711731, |
| "learning_rate": 4.168846203318613e-05, |
| "loss": 3.2583, |
| "step": 350000 |
| }, |
| { |
| "epoch": 14.753614635585718, |
| "eval_accuracy": 0.7454166410359169, |
| "eval_loss": 3.0359745025634766, |
| "eval_runtime": 189.972, |
| "eval_samples_per_second": 635.389, |
| "eval_steps_per_second": 6.622, |
| "step": 350000 |
| }, |
| { |
| "epoch": 14.77469122792227, |
| "grad_norm": 1.1582976579666138, |
| "learning_rate": 4.160270977147022e-05, |
| "loss": 3.257, |
| "step": 350500 |
| }, |
| { |
| "epoch": 14.79576782025882, |
| "grad_norm": 1.169987678527832, |
| "learning_rate": 4.1516957509754326e-05, |
| "loss": 3.2591, |
| "step": 351000 |
| }, |
| { |
| "epoch": 14.81684441259537, |
| "grad_norm": 1.1465296745300293, |
| "learning_rate": 4.143120524803842e-05, |
| "loss": 3.2543, |
| "step": 351500 |
| }, |
| { |
| "epoch": 14.837921004931923, |
| "grad_norm": 1.167558193206787, |
| "learning_rate": 4.1345452986322516e-05, |
| "loss": 3.2567, |
| "step": 352000 |
| }, |
| { |
| "epoch": 14.858997597268473, |
| "grad_norm": 1.1549878120422363, |
| "learning_rate": 4.1259700724606614e-05, |
| "loss": 3.2557, |
| "step": 352500 |
| }, |
| { |
| "epoch": 14.880074189605025, |
| "grad_norm": 1.2039144039154053, |
| "learning_rate": 4.117394846289071e-05, |
| "loss": 3.2542, |
| "step": 353000 |
| }, |
| { |
| "epoch": 14.901150781941576, |
| "grad_norm": 1.1518113613128662, |
| "learning_rate": 4.1088196201174804e-05, |
| "loss": 3.2569, |
| "step": 353500 |
| }, |
| { |
| "epoch": 14.922227374278126, |
| "grad_norm": 1.2069610357284546, |
| "learning_rate": 4.100244393945891e-05, |
| "loss": 3.2557, |
| "step": 354000 |
| }, |
| { |
| "epoch": 14.943303966614678, |
| "grad_norm": 1.1862930059432983, |
| "learning_rate": 4.0916691677743e-05, |
| "loss": 3.2605, |
| "step": 354500 |
| }, |
| { |
| "epoch": 14.964380558951229, |
| "grad_norm": 1.124613642692566, |
| "learning_rate": 4.08309394160271e-05, |
| "loss": 3.2575, |
| "step": 355000 |
| }, |
| { |
| "epoch": 14.985457151287779, |
| "grad_norm": 1.1725938320159912, |
| "learning_rate": 4.07451871543112e-05, |
| "loss": 3.2549, |
| "step": 355500 |
| }, |
| { |
| "epoch": 15.006533743624331, |
| "grad_norm": 1.1633851528167725, |
| "learning_rate": 4.0659434892595295e-05, |
| "loss": 3.251, |
| "step": 356000 |
| }, |
| { |
| "epoch": 15.027610335960881, |
| "grad_norm": 1.174697995185852, |
| "learning_rate": 4.057368263087939e-05, |
| "loss": 3.2466, |
| "step": 356500 |
| }, |
| { |
| "epoch": 15.048686928297434, |
| "grad_norm": 1.1682016849517822, |
| "learning_rate": 4.048793036916349e-05, |
| "loss": 3.2511, |
| "step": 357000 |
| }, |
| { |
| "epoch": 15.069763520633984, |
| "grad_norm": 1.2095342874526978, |
| "learning_rate": 4.0402178107447584e-05, |
| "loss": 3.2525, |
| "step": 357500 |
| }, |
| { |
| "epoch": 15.090840112970534, |
| "grad_norm": 1.1989741325378418, |
| "learning_rate": 4.031642584573168e-05, |
| "loss": 3.2518, |
| "step": 358000 |
| }, |
| { |
| "epoch": 15.111916705307086, |
| "grad_norm": 1.2167571783065796, |
| "learning_rate": 4.023067358401578e-05, |
| "loss": 3.2475, |
| "step": 358500 |
| }, |
| { |
| "epoch": 15.132993297643637, |
| "grad_norm": 1.197138786315918, |
| "learning_rate": 4.014492132229988e-05, |
| "loss": 3.2521, |
| "step": 359000 |
| }, |
| { |
| "epoch": 15.154069889980189, |
| "grad_norm": 1.197675347328186, |
| "learning_rate": 4.005916906058397e-05, |
| "loss": 3.2499, |
| "step": 359500 |
| }, |
| { |
| "epoch": 15.17514648231674, |
| "grad_norm": 1.173689365386963, |
| "learning_rate": 3.9973416798868075e-05, |
| "loss": 3.2508, |
| "step": 360000 |
| }, |
| { |
| "epoch": 15.17514648231674, |
| "eval_accuracy": 0.746121772811612, |
| "eval_loss": 3.032679319381714, |
| "eval_runtime": 189.81, |
| "eval_samples_per_second": 635.931, |
| "eval_steps_per_second": 6.628, |
| "step": 360000 |
| }, |
| { |
| "epoch": 15.19622307465329, |
| "grad_norm": 1.2497256994247437, |
| "learning_rate": 3.988766453715217e-05, |
| "loss": 3.2527, |
| "step": 360500 |
| }, |
| { |
| "epoch": 15.217299666989842, |
| "grad_norm": 1.183579444885254, |
| "learning_rate": 3.9801912275436265e-05, |
| "loss": 3.2509, |
| "step": 361000 |
| }, |
| { |
| "epoch": 15.238376259326392, |
| "grad_norm": 1.1390483379364014, |
| "learning_rate": 3.971616001372036e-05, |
| "loss": 3.2503, |
| "step": 361500 |
| }, |
| { |
| "epoch": 15.259452851662942, |
| "grad_norm": 1.193143606185913, |
| "learning_rate": 3.963040775200446e-05, |
| "loss": 3.2489, |
| "step": 362000 |
| }, |
| { |
| "epoch": 15.280529443999495, |
| "grad_norm": 1.2069480419158936, |
| "learning_rate": 3.954465549028856e-05, |
| "loss": 3.2483, |
| "step": 362500 |
| }, |
| { |
| "epoch": 15.301606036336045, |
| "grad_norm": 1.1501591205596924, |
| "learning_rate": 3.945890322857266e-05, |
| "loss": 3.2482, |
| "step": 363000 |
| }, |
| { |
| "epoch": 15.322682628672597, |
| "grad_norm": 1.154670000076294, |
| "learning_rate": 3.9373150966856756e-05, |
| "loss": 3.2456, |
| "step": 363500 |
| }, |
| { |
| "epoch": 15.343759221009147, |
| "grad_norm": 1.196118950843811, |
| "learning_rate": 3.928739870514085e-05, |
| "loss": 3.2479, |
| "step": 364000 |
| }, |
| { |
| "epoch": 15.364835813345698, |
| "grad_norm": 1.2030022144317627, |
| "learning_rate": 3.920164644342495e-05, |
| "loss": 3.2456, |
| "step": 364500 |
| }, |
| { |
| "epoch": 15.38591240568225, |
| "grad_norm": 1.1576471328735352, |
| "learning_rate": 3.9115894181709045e-05, |
| "loss": 3.2473, |
| "step": 365000 |
| }, |
| { |
| "epoch": 15.4069889980188, |
| "grad_norm": 1.1477421522140503, |
| "learning_rate": 3.903014191999314e-05, |
| "loss": 3.2449, |
| "step": 365500 |
| }, |
| { |
| "epoch": 15.42806559035535, |
| "grad_norm": 1.2031517028808594, |
| "learning_rate": 3.894438965827724e-05, |
| "loss": 3.2484, |
| "step": 366000 |
| }, |
| { |
| "epoch": 15.449142182691903, |
| "grad_norm": 1.1791483163833618, |
| "learning_rate": 3.885863739656134e-05, |
| "loss": 3.2459, |
| "step": 366500 |
| }, |
| { |
| "epoch": 15.470218775028453, |
| "grad_norm": 1.166008472442627, |
| "learning_rate": 3.877288513484543e-05, |
| "loss": 3.2452, |
| "step": 367000 |
| }, |
| { |
| "epoch": 15.491295367365005, |
| "grad_norm": 1.2359707355499268, |
| "learning_rate": 3.868713287312953e-05, |
| "loss": 3.2476, |
| "step": 367500 |
| }, |
| { |
| "epoch": 15.512371959701555, |
| "grad_norm": 1.191405177116394, |
| "learning_rate": 3.860138061141363e-05, |
| "loss": 3.2463, |
| "step": 368000 |
| }, |
| { |
| "epoch": 15.533448552038106, |
| "grad_norm": 1.1859800815582275, |
| "learning_rate": 3.8515628349697726e-05, |
| "loss": 3.2482, |
| "step": 368500 |
| }, |
| { |
| "epoch": 15.554525144374658, |
| "grad_norm": 1.2234588861465454, |
| "learning_rate": 3.842987608798182e-05, |
| "loss": 3.2447, |
| "step": 369000 |
| }, |
| { |
| "epoch": 15.575601736711208, |
| "grad_norm": 1.2031893730163574, |
| "learning_rate": 3.834412382626592e-05, |
| "loss": 3.2422, |
| "step": 369500 |
| }, |
| { |
| "epoch": 15.596678329047759, |
| "grad_norm": 1.1681305170059204, |
| "learning_rate": 3.8258371564550014e-05, |
| "loss": 3.2454, |
| "step": 370000 |
| }, |
| { |
| "epoch": 15.596678329047759, |
| "eval_accuracy": 0.7474133871608944, |
| "eval_loss": 3.0231688022613525, |
| "eval_runtime": 190.2315, |
| "eval_samples_per_second": 634.522, |
| "eval_steps_per_second": 6.613, |
| "step": 370000 |
| }, |
| { |
| "epoch": 15.61775492138431, |
| "grad_norm": 1.1642996072769165, |
| "learning_rate": 3.817261930283411e-05, |
| "loss": 3.2424, |
| "step": 370500 |
| }, |
| { |
| "epoch": 15.638831513720861, |
| "grad_norm": 1.2243826389312744, |
| "learning_rate": 3.808686704111821e-05, |
| "loss": 3.2449, |
| "step": 371000 |
| }, |
| { |
| "epoch": 15.659908106057413, |
| "grad_norm": 1.225684404373169, |
| "learning_rate": 3.800111477940231e-05, |
| "loss": 3.244, |
| "step": 371500 |
| }, |
| { |
| "epoch": 15.680984698393964, |
| "grad_norm": 1.191697359085083, |
| "learning_rate": 3.79153625176864e-05, |
| "loss": 3.2449, |
| "step": 372000 |
| }, |
| { |
| "epoch": 15.702061290730514, |
| "grad_norm": 1.1605284214019775, |
| "learning_rate": 3.7829610255970506e-05, |
| "loss": 3.2461, |
| "step": 372500 |
| }, |
| { |
| "epoch": 15.723137883067066, |
| "grad_norm": 1.1805402040481567, |
| "learning_rate": 3.77438579942546e-05, |
| "loss": 3.241, |
| "step": 373000 |
| }, |
| { |
| "epoch": 15.744214475403616, |
| "grad_norm": 1.2095226049423218, |
| "learning_rate": 3.7658105732538696e-05, |
| "loss": 3.2394, |
| "step": 373500 |
| }, |
| { |
| "epoch": 15.765291067740169, |
| "grad_norm": 1.1622179746627808, |
| "learning_rate": 3.7572353470822794e-05, |
| "loss": 3.2431, |
| "step": 374000 |
| }, |
| { |
| "epoch": 15.786367660076719, |
| "grad_norm": 1.2466975450515747, |
| "learning_rate": 3.748660120910689e-05, |
| "loss": 3.2383, |
| "step": 374500 |
| }, |
| { |
| "epoch": 15.80744425241327, |
| "grad_norm": 1.1983718872070312, |
| "learning_rate": 3.740084894739099e-05, |
| "loss": 3.2376, |
| "step": 375000 |
| }, |
| { |
| "epoch": 15.828520844749821, |
| "grad_norm": 1.159264326095581, |
| "learning_rate": 3.731509668567509e-05, |
| "loss": 3.2402, |
| "step": 375500 |
| }, |
| { |
| "epoch": 15.849597437086372, |
| "grad_norm": 1.1590213775634766, |
| "learning_rate": 3.722934442395919e-05, |
| "loss": 3.2438, |
| "step": 376000 |
| }, |
| { |
| "epoch": 15.870674029422922, |
| "grad_norm": 1.1751127243041992, |
| "learning_rate": 3.714359216224328e-05, |
| "loss": 3.2412, |
| "step": 376500 |
| }, |
| { |
| "epoch": 15.891750621759474, |
| "grad_norm": 1.1717164516448975, |
| "learning_rate": 3.705783990052738e-05, |
| "loss": 3.2437, |
| "step": 377000 |
| }, |
| { |
| "epoch": 15.912827214096025, |
| "grad_norm": 1.186503291130066, |
| "learning_rate": 3.6972087638811475e-05, |
| "loss": 3.2392, |
| "step": 377500 |
| }, |
| { |
| "epoch": 15.933903806432577, |
| "grad_norm": 1.137797474861145, |
| "learning_rate": 3.6886335377095574e-05, |
| "loss": 3.241, |
| "step": 378000 |
| }, |
| { |
| "epoch": 15.954980398769127, |
| "grad_norm": 1.1282835006713867, |
| "learning_rate": 3.680058311537967e-05, |
| "loss": 3.2402, |
| "step": 378500 |
| }, |
| { |
| "epoch": 15.976056991105677, |
| "grad_norm": 1.2031331062316895, |
| "learning_rate": 3.671483085366377e-05, |
| "loss": 3.2417, |
| "step": 379000 |
| }, |
| { |
| "epoch": 15.99713358344223, |
| "grad_norm": 1.207740068435669, |
| "learning_rate": 3.662907859194786e-05, |
| "loss": 3.2395, |
| "step": 379500 |
| }, |
| { |
| "epoch": 16.01821017577878, |
| "grad_norm": 1.2053711414337158, |
| "learning_rate": 3.654332633023197e-05, |
| "loss": 3.2386, |
| "step": 380000 |
| }, |
| { |
| "epoch": 16.01821017577878, |
| "eval_accuracy": 0.748361154953501, |
| "eval_loss": 3.018552780151367, |
| "eval_runtime": 187.7794, |
| "eval_samples_per_second": 642.807, |
| "eval_steps_per_second": 6.699, |
| "step": 380000 |
| }, |
| { |
| "epoch": 16.039286768115332, |
| "grad_norm": 1.1941330432891846, |
| "learning_rate": 3.645757406851606e-05, |
| "loss": 3.2361, |
| "step": 380500 |
| }, |
| { |
| "epoch": 16.06036336045188, |
| "grad_norm": 1.2262500524520874, |
| "learning_rate": 3.6371821806800157e-05, |
| "loss": 3.2343, |
| "step": 381000 |
| }, |
| { |
| "epoch": 16.081439952788433, |
| "grad_norm": 1.1395760774612427, |
| "learning_rate": 3.6286069545084255e-05, |
| "loss": 3.2382, |
| "step": 381500 |
| }, |
| { |
| "epoch": 16.102516545124985, |
| "grad_norm": 1.1749221086502075, |
| "learning_rate": 3.620031728336835e-05, |
| "loss": 3.2359, |
| "step": 382000 |
| }, |
| { |
| "epoch": 16.123593137461537, |
| "grad_norm": 1.1869754791259766, |
| "learning_rate": 3.6114565021652445e-05, |
| "loss": 3.2296, |
| "step": 382500 |
| }, |
| { |
| "epoch": 16.144669729798085, |
| "grad_norm": 1.2069379091262817, |
| "learning_rate": 3.602881275993655e-05, |
| "loss": 3.2328, |
| "step": 383000 |
| }, |
| { |
| "epoch": 16.165746322134638, |
| "grad_norm": 1.226843237876892, |
| "learning_rate": 3.594306049822064e-05, |
| "loss": 3.2381, |
| "step": 383500 |
| }, |
| { |
| "epoch": 16.18682291447119, |
| "grad_norm": 1.2125753164291382, |
| "learning_rate": 3.585730823650474e-05, |
| "loss": 3.2349, |
| "step": 384000 |
| }, |
| { |
| "epoch": 16.20789950680774, |
| "grad_norm": 1.2056090831756592, |
| "learning_rate": 3.577155597478883e-05, |
| "loss": 3.236, |
| "step": 384500 |
| }, |
| { |
| "epoch": 16.22897609914429, |
| "grad_norm": 1.1926831007003784, |
| "learning_rate": 3.5685803713072936e-05, |
| "loss": 3.2305, |
| "step": 385000 |
| }, |
| { |
| "epoch": 16.250052691480843, |
| "grad_norm": 1.1919599771499634, |
| "learning_rate": 3.560005145135703e-05, |
| "loss": 3.2336, |
| "step": 385500 |
| }, |
| { |
| "epoch": 16.27112928381739, |
| "grad_norm": 1.2114341259002686, |
| "learning_rate": 3.5514299189641126e-05, |
| "loss": 3.2305, |
| "step": 386000 |
| }, |
| { |
| "epoch": 16.292205876153943, |
| "grad_norm": 1.1763794422149658, |
| "learning_rate": 3.5428546927925224e-05, |
| "loss": 3.2351, |
| "step": 386500 |
| }, |
| { |
| "epoch": 16.313282468490495, |
| "grad_norm": 1.1750164031982422, |
| "learning_rate": 3.534279466620932e-05, |
| "loss": 3.2303, |
| "step": 387000 |
| }, |
| { |
| "epoch": 16.334359060827044, |
| "grad_norm": 1.1961358785629272, |
| "learning_rate": 3.525704240449342e-05, |
| "loss": 3.2323, |
| "step": 387500 |
| }, |
| { |
| "epoch": 16.355435653163596, |
| "grad_norm": 1.1813045740127563, |
| "learning_rate": 3.517129014277752e-05, |
| "loss": 3.2299, |
| "step": 388000 |
| }, |
| { |
| "epoch": 16.376512245500148, |
| "grad_norm": 1.1918658018112183, |
| "learning_rate": 3.508553788106161e-05, |
| "loss": 3.2309, |
| "step": 388500 |
| }, |
| { |
| "epoch": 16.3975888378367, |
| "grad_norm": 1.1767683029174805, |
| "learning_rate": 3.499978561934571e-05, |
| "loss": 3.2285, |
| "step": 389000 |
| }, |
| { |
| "epoch": 16.41866543017325, |
| "grad_norm": 1.125380516052246, |
| "learning_rate": 3.491403335762981e-05, |
| "loss": 3.2308, |
| "step": 389500 |
| }, |
| { |
| "epoch": 16.4397420225098, |
| "grad_norm": 1.211596965789795, |
| "learning_rate": 3.4828281095913906e-05, |
| "loss": 3.2334, |
| "step": 390000 |
| }, |
| { |
| "epoch": 16.4397420225098, |
| "eval_accuracy": 0.7491886340780046, |
| "eval_loss": 3.0146641731262207, |
| "eval_runtime": 193.4448, |
| "eval_samples_per_second": 623.982, |
| "eval_steps_per_second": 6.503, |
| "step": 390000 |
| }, |
| { |
| "epoch": 16.460818614846353, |
| "grad_norm": 1.236082673072815, |
| "learning_rate": 3.4742528834198004e-05, |
| "loss": 3.2298, |
| "step": 390500 |
| }, |
| { |
| "epoch": 16.4818952071829, |
| "grad_norm": 1.1778305768966675, |
| "learning_rate": 3.46567765724821e-05, |
| "loss": 3.2332, |
| "step": 391000 |
| }, |
| { |
| "epoch": 16.502971799519454, |
| "grad_norm": 1.19925856590271, |
| "learning_rate": 3.45710243107662e-05, |
| "loss": 3.2314, |
| "step": 391500 |
| }, |
| { |
| "epoch": 16.524048391856006, |
| "grad_norm": 1.1969057321548462, |
| "learning_rate": 3.448527204905029e-05, |
| "loss": 3.2346, |
| "step": 392000 |
| }, |
| { |
| "epoch": 16.545124984192555, |
| "grad_norm": 1.201365351676941, |
| "learning_rate": 3.43995197873344e-05, |
| "loss": 3.2283, |
| "step": 392500 |
| }, |
| { |
| "epoch": 16.566201576529107, |
| "grad_norm": 1.212031602859497, |
| "learning_rate": 3.431376752561849e-05, |
| "loss": 3.2307, |
| "step": 393000 |
| }, |
| { |
| "epoch": 16.58727816886566, |
| "grad_norm": 1.1186423301696777, |
| "learning_rate": 3.422801526390259e-05, |
| "loss": 3.2325, |
| "step": 393500 |
| }, |
| { |
| "epoch": 16.608354761202207, |
| "grad_norm": 1.1851462125778198, |
| "learning_rate": 3.4142263002186686e-05, |
| "loss": 3.2314, |
| "step": 394000 |
| }, |
| { |
| "epoch": 16.62943135353876, |
| "grad_norm": 1.180517554283142, |
| "learning_rate": 3.4056510740470784e-05, |
| "loss": 3.2337, |
| "step": 394500 |
| }, |
| { |
| "epoch": 16.65050794587531, |
| "grad_norm": 1.208378791809082, |
| "learning_rate": 3.3970758478754875e-05, |
| "loss": 3.23, |
| "step": 395000 |
| }, |
| { |
| "epoch": 16.67158453821186, |
| "grad_norm": 1.177207589149475, |
| "learning_rate": 3.388500621703898e-05, |
| "loss": 3.2302, |
| "step": 395500 |
| }, |
| { |
| "epoch": 16.692661130548412, |
| "grad_norm": 1.2077230215072632, |
| "learning_rate": 3.379925395532307e-05, |
| "loss": 3.2311, |
| "step": 396000 |
| }, |
| { |
| "epoch": 16.713737722884964, |
| "grad_norm": 1.1417843103408813, |
| "learning_rate": 3.371350169360717e-05, |
| "loss": 3.2278, |
| "step": 396500 |
| }, |
| { |
| "epoch": 16.734814315221517, |
| "grad_norm": 1.1677274703979492, |
| "learning_rate": 3.362774943189127e-05, |
| "loss": 3.232, |
| "step": 397000 |
| }, |
| { |
| "epoch": 16.755890907558065, |
| "grad_norm": 1.1267443895339966, |
| "learning_rate": 3.354199717017537e-05, |
| "loss": 3.2281, |
| "step": 397500 |
| }, |
| { |
| "epoch": 16.776967499894617, |
| "grad_norm": 1.1946439743041992, |
| "learning_rate": 3.345624490845946e-05, |
| "loss": 3.2276, |
| "step": 398000 |
| }, |
| { |
| "epoch": 16.79804409223117, |
| "grad_norm": 1.1730728149414062, |
| "learning_rate": 3.3370492646743564e-05, |
| "loss": 3.2268, |
| "step": 398500 |
| }, |
| { |
| "epoch": 16.819120684567718, |
| "grad_norm": 1.1747136116027832, |
| "learning_rate": 3.3284740385027655e-05, |
| "loss": 3.2328, |
| "step": 399000 |
| }, |
| { |
| "epoch": 16.84019727690427, |
| "grad_norm": 1.1446260213851929, |
| "learning_rate": 3.319898812331175e-05, |
| "loss": 3.2243, |
| "step": 399500 |
| }, |
| { |
| "epoch": 16.861273869240822, |
| "grad_norm": 1.19792902469635, |
| "learning_rate": 3.311323586159585e-05, |
| "loss": 3.2251, |
| "step": 400000 |
| }, |
| { |
| "epoch": 16.861273869240822, |
| "eval_accuracy": 0.7499386341296485, |
| "eval_loss": 3.010554552078247, |
| "eval_runtime": 190.8635, |
| "eval_samples_per_second": 632.421, |
| "eval_steps_per_second": 6.591, |
| "step": 400000 |
| }, |
| { |
| "epoch": 16.88235046157737, |
| "grad_norm": 1.2194592952728271, |
| "learning_rate": 3.302748359987995e-05, |
| "loss": 3.2248, |
| "step": 400500 |
| }, |
| { |
| "epoch": 16.903427053913923, |
| "grad_norm": 1.191601037979126, |
| "learning_rate": 3.294173133816404e-05, |
| "loss": 3.2263, |
| "step": 401000 |
| }, |
| { |
| "epoch": 16.924503646250475, |
| "grad_norm": 1.2041834592819214, |
| "learning_rate": 3.2855979076448147e-05, |
| "loss": 3.2302, |
| "step": 401500 |
| }, |
| { |
| "epoch": 16.945580238587024, |
| "grad_norm": 1.163550615310669, |
| "learning_rate": 3.277022681473224e-05, |
| "loss": 3.2278, |
| "step": 402000 |
| }, |
| { |
| "epoch": 16.966656830923576, |
| "grad_norm": 1.1868038177490234, |
| "learning_rate": 3.2684474553016336e-05, |
| "loss": 3.2266, |
| "step": 402500 |
| }, |
| { |
| "epoch": 16.987733423260128, |
| "grad_norm": 1.1726408004760742, |
| "learning_rate": 3.2598722291300435e-05, |
| "loss": 3.228, |
| "step": 403000 |
| }, |
| { |
| "epoch": 17.00881001559668, |
| "grad_norm": 1.2119767665863037, |
| "learning_rate": 3.251297002958453e-05, |
| "loss": 3.226, |
| "step": 403500 |
| }, |
| { |
| "epoch": 17.02988660793323, |
| "grad_norm": 1.2066272497177124, |
| "learning_rate": 3.242721776786863e-05, |
| "loss": 3.2482, |
| "step": 404000 |
| }, |
| { |
| "epoch": 17.05096320026978, |
| "grad_norm": 1.1788074970245361, |
| "learning_rate": 3.234146550615272e-05, |
| "loss": 3.244, |
| "step": 404500 |
| }, |
| { |
| "epoch": 17.072039792606333, |
| "grad_norm": 1.258504033088684, |
| "learning_rate": 3.225571324443683e-05, |
| "loss": 3.2443, |
| "step": 405000 |
| }, |
| { |
| "epoch": 17.09311638494288, |
| "grad_norm": 1.2046279907226562, |
| "learning_rate": 3.216996098272092e-05, |
| "loss": 3.2351, |
| "step": 405500 |
| }, |
| { |
| "epoch": 17.114192977279433, |
| "grad_norm": 1.1949303150177002, |
| "learning_rate": 3.208420872100502e-05, |
| "loss": 3.2236, |
| "step": 406000 |
| }, |
| { |
| "epoch": 17.135269569615986, |
| "grad_norm": 1.1790212392807007, |
| "learning_rate": 3.1998456459289116e-05, |
| "loss": 3.219, |
| "step": 406500 |
| }, |
| { |
| "epoch": 17.156346161952534, |
| "grad_norm": 1.250762701034546, |
| "learning_rate": 3.1912704197573214e-05, |
| "loss": 3.2196, |
| "step": 407000 |
| }, |
| { |
| "epoch": 17.177422754289086, |
| "grad_norm": 1.2282357215881348, |
| "learning_rate": 3.1826951935857306e-05, |
| "loss": 3.2187, |
| "step": 407500 |
| }, |
| { |
| "epoch": 17.19849934662564, |
| "grad_norm": 1.1928554773330688, |
| "learning_rate": 3.174119967414141e-05, |
| "loss": 3.2232, |
| "step": 408000 |
| }, |
| { |
| "epoch": 17.219575938962187, |
| "grad_norm": 1.2214782238006592, |
| "learning_rate": 3.16554474124255e-05, |
| "loss": 3.22, |
| "step": 408500 |
| }, |
| { |
| "epoch": 17.24065253129874, |
| "grad_norm": 1.195587158203125, |
| "learning_rate": 3.15696951507096e-05, |
| "loss": 3.223, |
| "step": 409000 |
| }, |
| { |
| "epoch": 17.26172912363529, |
| "grad_norm": 1.2169852256774902, |
| "learning_rate": 3.14839428889937e-05, |
| "loss": 3.2207, |
| "step": 409500 |
| }, |
| { |
| "epoch": 17.282805715971843, |
| "grad_norm": 1.2571474313735962, |
| "learning_rate": 3.13981906272778e-05, |
| "loss": 3.2228, |
| "step": 410000 |
| }, |
| { |
| "epoch": 17.282805715971843, |
| "eval_accuracy": 0.7510824818396372, |
| "eval_loss": 3.0034828186035156, |
| "eval_runtime": 189.8633, |
| "eval_samples_per_second": 635.752, |
| "eval_steps_per_second": 6.626, |
| "step": 410000 |
| }, |
| { |
| "epoch": 17.303882308308392, |
| "grad_norm": 1.2489392757415771, |
| "learning_rate": 3.131243836556189e-05, |
| "loss": 3.2241, |
| "step": 410500 |
| }, |
| { |
| "epoch": 17.324958900644944, |
| "grad_norm": 1.1909772157669067, |
| "learning_rate": 3.1226686103845994e-05, |
| "loss": 3.2179, |
| "step": 411000 |
| }, |
| { |
| "epoch": 17.346035492981496, |
| "grad_norm": 1.2130558490753174, |
| "learning_rate": 3.1140933842130086e-05, |
| "loss": 3.2185, |
| "step": 411500 |
| }, |
| { |
| "epoch": 17.367112085318045, |
| "grad_norm": 1.1711021661758423, |
| "learning_rate": 3.1055181580414184e-05, |
| "loss": 3.2216, |
| "step": 412000 |
| }, |
| { |
| "epoch": 17.388188677654597, |
| "grad_norm": 1.2047263383865356, |
| "learning_rate": 3.096942931869828e-05, |
| "loss": 3.2229, |
| "step": 412500 |
| }, |
| { |
| "epoch": 17.40926526999115, |
| "grad_norm": 1.1802482604980469, |
| "learning_rate": 3.088367705698238e-05, |
| "loss": 3.2163, |
| "step": 413000 |
| }, |
| { |
| "epoch": 17.430341862327698, |
| "grad_norm": 1.137651801109314, |
| "learning_rate": 3.079792479526647e-05, |
| "loss": 3.2195, |
| "step": 413500 |
| }, |
| { |
| "epoch": 17.45141845466425, |
| "grad_norm": 1.1307164430618286, |
| "learning_rate": 3.071217253355058e-05, |
| "loss": 3.2173, |
| "step": 414000 |
| }, |
| { |
| "epoch": 17.472495047000802, |
| "grad_norm": 1.190506100654602, |
| "learning_rate": 3.062642027183467e-05, |
| "loss": 3.219, |
| "step": 414500 |
| }, |
| { |
| "epoch": 17.49357163933735, |
| "grad_norm": 1.1782413721084595, |
| "learning_rate": 3.054066801011877e-05, |
| "loss": 3.2178, |
| "step": 415000 |
| }, |
| { |
| "epoch": 17.514648231673903, |
| "grad_norm": 1.1930475234985352, |
| "learning_rate": 3.045491574840287e-05, |
| "loss": 3.2205, |
| "step": 415500 |
| }, |
| { |
| "epoch": 17.535724824010455, |
| "grad_norm": 1.2135998010635376, |
| "learning_rate": 3.0369163486686964e-05, |
| "loss": 3.2162, |
| "step": 416000 |
| }, |
| { |
| "epoch": 17.556801416347007, |
| "grad_norm": 1.2325297594070435, |
| "learning_rate": 3.028341122497106e-05, |
| "loss": 3.2158, |
| "step": 416500 |
| }, |
| { |
| "epoch": 17.577878008683555, |
| "grad_norm": 1.2288234233856201, |
| "learning_rate": 3.019765896325516e-05, |
| "loss": 3.2201, |
| "step": 417000 |
| }, |
| { |
| "epoch": 17.598954601020107, |
| "grad_norm": 1.1923421621322632, |
| "learning_rate": 3.0111906701539255e-05, |
| "loss": 3.2192, |
| "step": 417500 |
| }, |
| { |
| "epoch": 17.62003119335666, |
| "grad_norm": 1.2019927501678467, |
| "learning_rate": 3.002615443982335e-05, |
| "loss": 3.2168, |
| "step": 418000 |
| }, |
| { |
| "epoch": 17.641107785693208, |
| "grad_norm": 1.2093620300292969, |
| "learning_rate": 2.9940402178107452e-05, |
| "loss": 3.2152, |
| "step": 418500 |
| }, |
| { |
| "epoch": 17.66218437802976, |
| "grad_norm": 1.2123472690582275, |
| "learning_rate": 2.9854649916391547e-05, |
| "loss": 3.2166, |
| "step": 419000 |
| }, |
| { |
| "epoch": 17.683260970366312, |
| "grad_norm": 1.2471553087234497, |
| "learning_rate": 2.976889765467564e-05, |
| "loss": 3.221, |
| "step": 419500 |
| }, |
| { |
| "epoch": 17.70433756270286, |
| "grad_norm": 1.254834532737732, |
| "learning_rate": 2.9683145392959737e-05, |
| "loss": 3.2162, |
| "step": 420000 |
| }, |
| { |
| "epoch": 17.70433756270286, |
| "eval_accuracy": 0.7516955630951956, |
| "eval_loss": 3.0003249645233154, |
| "eval_runtime": 191.709, |
| "eval_samples_per_second": 629.631, |
| "eval_steps_per_second": 6.562, |
| "step": 420000 |
| }, |
| { |
| "epoch": 17.725414155039413, |
| "grad_norm": 1.231488823890686, |
| "learning_rate": 2.9597393131243838e-05, |
| "loss": 3.2155, |
| "step": 420500 |
| }, |
| { |
| "epoch": 17.746490747375965, |
| "grad_norm": 1.2041956186294556, |
| "learning_rate": 2.9511640869527933e-05, |
| "loss": 3.2139, |
| "step": 421000 |
| }, |
| { |
| "epoch": 17.767567339712514, |
| "grad_norm": 1.2090623378753662, |
| "learning_rate": 2.942588860781203e-05, |
| "loss": 3.2135, |
| "step": 421500 |
| }, |
| { |
| "epoch": 17.788643932049066, |
| "grad_norm": 1.2198864221572876, |
| "learning_rate": 2.934013634609613e-05, |
| "loss": 3.216, |
| "step": 422000 |
| }, |
| { |
| "epoch": 17.809720524385618, |
| "grad_norm": 1.1801376342773438, |
| "learning_rate": 2.9254384084380225e-05, |
| "loss": 3.217, |
| "step": 422500 |
| }, |
| { |
| "epoch": 17.830797116722167, |
| "grad_norm": 1.2321057319641113, |
| "learning_rate": 2.9168631822664323e-05, |
| "loss": 3.2166, |
| "step": 423000 |
| }, |
| { |
| "epoch": 17.85187370905872, |
| "grad_norm": 1.1681984663009644, |
| "learning_rate": 2.908287956094842e-05, |
| "loss": 3.2156, |
| "step": 423500 |
| }, |
| { |
| "epoch": 17.87295030139527, |
| "grad_norm": 1.2125357389450073, |
| "learning_rate": 2.899712729923252e-05, |
| "loss": 3.2146, |
| "step": 424000 |
| }, |
| { |
| "epoch": 17.894026893731823, |
| "grad_norm": 1.1644959449768066, |
| "learning_rate": 2.8911375037516615e-05, |
| "loss": 3.2136, |
| "step": 424500 |
| }, |
| { |
| "epoch": 17.91510348606837, |
| "grad_norm": 1.2323737144470215, |
| "learning_rate": 2.8825622775800716e-05, |
| "loss": 3.217, |
| "step": 425000 |
| }, |
| { |
| "epoch": 17.936180078404924, |
| "grad_norm": 1.2196662425994873, |
| "learning_rate": 2.873987051408481e-05, |
| "loss": 3.2145, |
| "step": 425500 |
| }, |
| { |
| "epoch": 17.957256670741476, |
| "grad_norm": 1.1980717182159424, |
| "learning_rate": 2.8654118252368906e-05, |
| "loss": 3.2123, |
| "step": 426000 |
| }, |
| { |
| "epoch": 17.978333263078024, |
| "grad_norm": 1.2101510763168335, |
| "learning_rate": 2.8568365990653008e-05, |
| "loss": 3.2114, |
| "step": 426500 |
| }, |
| { |
| "epoch": 17.999409855414576, |
| "grad_norm": 1.1909844875335693, |
| "learning_rate": 2.8482613728937103e-05, |
| "loss": 3.2154, |
| "step": 427000 |
| }, |
| { |
| "epoch": 18.02048644775113, |
| "grad_norm": 1.2181487083435059, |
| "learning_rate": 2.8396861467221198e-05, |
| "loss": 3.212, |
| "step": 427500 |
| }, |
| { |
| "epoch": 18.041563040087677, |
| "grad_norm": 1.216494083404541, |
| "learning_rate": 2.83111092055053e-05, |
| "loss": 3.208, |
| "step": 428000 |
| }, |
| { |
| "epoch": 18.06263963242423, |
| "grad_norm": 1.199052333831787, |
| "learning_rate": 2.8225356943789394e-05, |
| "loss": 3.2095, |
| "step": 428500 |
| }, |
| { |
| "epoch": 18.08371622476078, |
| "grad_norm": 1.2267423868179321, |
| "learning_rate": 2.813960468207349e-05, |
| "loss": 3.2085, |
| "step": 429000 |
| }, |
| { |
| "epoch": 18.10479281709733, |
| "grad_norm": 1.1572250127792358, |
| "learning_rate": 2.805385242035759e-05, |
| "loss": 3.208, |
| "step": 429500 |
| }, |
| { |
| "epoch": 18.125869409433882, |
| "grad_norm": 1.2807753086090088, |
| "learning_rate": 2.7968100158641686e-05, |
| "loss": 3.2079, |
| "step": 430000 |
| }, |
| { |
| "epoch": 18.125869409433882, |
| "eval_accuracy": 0.7523756487337883, |
| "eval_loss": 2.9971184730529785, |
| "eval_runtime": 190.304, |
| "eval_samples_per_second": 634.28, |
| "eval_steps_per_second": 6.61, |
| "step": 430000 |
| }, |
| { |
| "epoch": 18.146946001770434, |
| "grad_norm": 1.2112754583358765, |
| "learning_rate": 2.788234789692578e-05, |
| "loss": 3.2098, |
| "step": 430500 |
| }, |
| { |
| "epoch": 18.168022594106986, |
| "grad_norm": 1.2492507696151733, |
| "learning_rate": 2.7796595635209882e-05, |
| "loss": 3.21, |
| "step": 431000 |
| }, |
| { |
| "epoch": 18.189099186443535, |
| "grad_norm": 1.2437224388122559, |
| "learning_rate": 2.7710843373493977e-05, |
| "loss": 3.2112, |
| "step": 431500 |
| }, |
| { |
| "epoch": 18.210175778780087, |
| "grad_norm": 1.2143077850341797, |
| "learning_rate": 2.7625091111778072e-05, |
| "loss": 3.214, |
| "step": 432000 |
| }, |
| { |
| "epoch": 18.23125237111664, |
| "grad_norm": 1.2276734113693237, |
| "learning_rate": 2.7539338850062174e-05, |
| "loss": 3.2088, |
| "step": 432500 |
| }, |
| { |
| "epoch": 18.252328963453188, |
| "grad_norm": 1.1820963621139526, |
| "learning_rate": 2.745358658834627e-05, |
| "loss": 3.2088, |
| "step": 433000 |
| }, |
| { |
| "epoch": 18.27340555578974, |
| "grad_norm": 1.200482964515686, |
| "learning_rate": 2.7367834326630364e-05, |
| "loss": 3.2076, |
| "step": 433500 |
| }, |
| { |
| "epoch": 18.294482148126292, |
| "grad_norm": 1.182140588760376, |
| "learning_rate": 2.7282082064914465e-05, |
| "loss": 3.2084, |
| "step": 434000 |
| }, |
| { |
| "epoch": 18.31555874046284, |
| "grad_norm": 1.2205382585525513, |
| "learning_rate": 2.719632980319856e-05, |
| "loss": 3.2061, |
| "step": 434500 |
| }, |
| { |
| "epoch": 18.336635332799393, |
| "grad_norm": 1.1830172538757324, |
| "learning_rate": 2.7110577541482655e-05, |
| "loss": 3.2076, |
| "step": 435000 |
| }, |
| { |
| "epoch": 18.357711925135945, |
| "grad_norm": 1.2060185670852661, |
| "learning_rate": 2.7024825279766757e-05, |
| "loss": 3.2093, |
| "step": 435500 |
| }, |
| { |
| "epoch": 18.378788517472493, |
| "grad_norm": 1.237815022468567, |
| "learning_rate": 2.6939073018050852e-05, |
| "loss": 3.2079, |
| "step": 436000 |
| }, |
| { |
| "epoch": 18.399865109809046, |
| "grad_norm": 1.1929911375045776, |
| "learning_rate": 2.685332075633495e-05, |
| "loss": 3.2089, |
| "step": 436500 |
| }, |
| { |
| "epoch": 18.420941702145598, |
| "grad_norm": 1.1721742153167725, |
| "learning_rate": 2.6767568494619045e-05, |
| "loss": 3.2087, |
| "step": 437000 |
| }, |
| { |
| "epoch": 18.44201829448215, |
| "grad_norm": 1.2209805250167847, |
| "learning_rate": 2.6681816232903143e-05, |
| "loss": 3.206, |
| "step": 437500 |
| }, |
| { |
| "epoch": 18.4630948868187, |
| "grad_norm": 1.2465707063674927, |
| "learning_rate": 2.6596063971187242e-05, |
| "loss": 3.2012, |
| "step": 438000 |
| }, |
| { |
| "epoch": 18.48417147915525, |
| "grad_norm": 1.258580207824707, |
| "learning_rate": 2.6510311709471337e-05, |
| "loss": 3.2034, |
| "step": 438500 |
| }, |
| { |
| "epoch": 18.505248071491803, |
| "grad_norm": 1.1957414150238037, |
| "learning_rate": 2.642455944775544e-05, |
| "loss": 3.2054, |
| "step": 439000 |
| }, |
| { |
| "epoch": 18.52632466382835, |
| "grad_norm": 1.1814430952072144, |
| "learning_rate": 2.6338807186039533e-05, |
| "loss": 3.2048, |
| "step": 439500 |
| }, |
| { |
| "epoch": 18.547401256164903, |
| "grad_norm": 1.2349215745925903, |
| "learning_rate": 2.6253054924323628e-05, |
| "loss": 3.2024, |
| "step": 440000 |
| }, |
| { |
| "epoch": 18.547401256164903, |
| "eval_accuracy": 0.753022914765621, |
| "eval_loss": 2.992814064025879, |
| "eval_runtime": 193.4818, |
| "eval_samples_per_second": 623.862, |
| "eval_steps_per_second": 6.502, |
| "step": 440000 |
| }, |
| { |
| "epoch": 18.568477848501455, |
| "grad_norm": 1.229929804801941, |
| "learning_rate": 2.616730266260773e-05, |
| "loss": 3.2031, |
| "step": 440500 |
| }, |
| { |
| "epoch": 18.589554440838004, |
| "grad_norm": 1.2515631914138794, |
| "learning_rate": 2.6081550400891825e-05, |
| "loss": 3.2047, |
| "step": 441000 |
| }, |
| { |
| "epoch": 18.610631033174556, |
| "grad_norm": 1.198481798171997, |
| "learning_rate": 2.599579813917592e-05, |
| "loss": 3.2037, |
| "step": 441500 |
| }, |
| { |
| "epoch": 18.63170762551111, |
| "grad_norm": 1.230067491531372, |
| "learning_rate": 2.591004587746002e-05, |
| "loss": 3.204, |
| "step": 442000 |
| }, |
| { |
| "epoch": 18.652784217847657, |
| "grad_norm": 1.2100332975387573, |
| "learning_rate": 2.5824293615744116e-05, |
| "loss": 3.2067, |
| "step": 442500 |
| }, |
| { |
| "epoch": 18.67386081018421, |
| "grad_norm": 1.2223870754241943, |
| "learning_rate": 2.573854135402821e-05, |
| "loss": 3.2086, |
| "step": 443000 |
| }, |
| { |
| "epoch": 18.69493740252076, |
| "grad_norm": 1.2364284992218018, |
| "learning_rate": 2.5652789092312313e-05, |
| "loss": 3.2046, |
| "step": 443500 |
| }, |
| { |
| "epoch": 18.716013994857313, |
| "grad_norm": 1.2450991868972778, |
| "learning_rate": 2.5567036830596408e-05, |
| "loss": 3.2036, |
| "step": 444000 |
| }, |
| { |
| "epoch": 18.737090587193862, |
| "grad_norm": 1.2010419368743896, |
| "learning_rate": 2.5481284568880503e-05, |
| "loss": 3.2057, |
| "step": 444500 |
| }, |
| { |
| "epoch": 18.758167179530414, |
| "grad_norm": 1.2111302614212036, |
| "learning_rate": 2.5395532307164605e-05, |
| "loss": 3.1994, |
| "step": 445000 |
| }, |
| { |
| "epoch": 18.779243771866966, |
| "grad_norm": 1.3117402791976929, |
| "learning_rate": 2.53097800454487e-05, |
| "loss": 3.2067, |
| "step": 445500 |
| }, |
| { |
| "epoch": 18.800320364203515, |
| "grad_norm": 1.2030795812606812, |
| "learning_rate": 2.5224027783732794e-05, |
| "loss": 3.2045, |
| "step": 446000 |
| }, |
| { |
| "epoch": 18.821396956540067, |
| "grad_norm": 1.2809398174285889, |
| "learning_rate": 2.5138275522016896e-05, |
| "loss": 3.1965, |
| "step": 446500 |
| }, |
| { |
| "epoch": 18.84247354887662, |
| "grad_norm": 1.2198524475097656, |
| "learning_rate": 2.505252326030099e-05, |
| "loss": 3.1995, |
| "step": 447000 |
| }, |
| { |
| "epoch": 18.863550141213167, |
| "grad_norm": 1.235028624534607, |
| "learning_rate": 2.496677099858509e-05, |
| "loss": 3.202, |
| "step": 447500 |
| }, |
| { |
| "epoch": 18.88462673354972, |
| "grad_norm": 1.2018718719482422, |
| "learning_rate": 2.4881018736869184e-05, |
| "loss": 3.204, |
| "step": 448000 |
| }, |
| { |
| "epoch": 18.90570332588627, |
| "grad_norm": 1.2342514991760254, |
| "learning_rate": 2.4795266475153283e-05, |
| "loss": 3.2007, |
| "step": 448500 |
| }, |
| { |
| "epoch": 18.92677991822282, |
| "grad_norm": 1.212725281715393, |
| "learning_rate": 2.470951421343738e-05, |
| "loss": 3.2048, |
| "step": 449000 |
| }, |
| { |
| "epoch": 18.947856510559372, |
| "grad_norm": 1.2071921825408936, |
| "learning_rate": 2.4623761951721476e-05, |
| "loss": 3.1997, |
| "step": 449500 |
| }, |
| { |
| "epoch": 18.968933102895924, |
| "grad_norm": 1.1980541944503784, |
| "learning_rate": 2.4538009690005574e-05, |
| "loss": 3.2014, |
| "step": 450000 |
| }, |
| { |
| "epoch": 18.968933102895924, |
| "eval_accuracy": 0.7540922315211985, |
| "eval_loss": 2.9856114387512207, |
| "eval_runtime": 182.4647, |
| "eval_samples_per_second": 661.531, |
| "eval_steps_per_second": 6.894, |
| "step": 450000 |
| }, |
| { |
| "epoch": 18.990009695232473, |
| "grad_norm": 1.2336691617965698, |
| "learning_rate": 2.4452257428289672e-05, |
| "loss": 3.1998, |
| "step": 450500 |
| }, |
| { |
| "epoch": 19.011086287569025, |
| "grad_norm": 1.2340065240859985, |
| "learning_rate": 2.436650516657377e-05, |
| "loss": 3.1989, |
| "step": 451000 |
| }, |
| { |
| "epoch": 19.032162879905577, |
| "grad_norm": 1.2417525053024292, |
| "learning_rate": 2.428075290485787e-05, |
| "loss": 3.197, |
| "step": 451500 |
| }, |
| { |
| "epoch": 19.05323947224213, |
| "grad_norm": 1.202978253364563, |
| "learning_rate": 2.4195000643141964e-05, |
| "loss": 3.1987, |
| "step": 452000 |
| }, |
| { |
| "epoch": 19.074316064578678, |
| "grad_norm": 1.2290364503860474, |
| "learning_rate": 2.4109248381426062e-05, |
| "loss": 3.2003, |
| "step": 452500 |
| }, |
| { |
| "epoch": 19.09539265691523, |
| "grad_norm": 1.206170916557312, |
| "learning_rate": 2.402349611971016e-05, |
| "loss": 3.2001, |
| "step": 453000 |
| }, |
| { |
| "epoch": 19.116469249251782, |
| "grad_norm": 1.145770788192749, |
| "learning_rate": 2.3937743857994255e-05, |
| "loss": 3.1954, |
| "step": 453500 |
| }, |
| { |
| "epoch": 19.13754584158833, |
| "grad_norm": 1.2028239965438843, |
| "learning_rate": 2.3851991596278354e-05, |
| "loss": 3.1977, |
| "step": 454000 |
| }, |
| { |
| "epoch": 19.158622433924883, |
| "grad_norm": 1.2532691955566406, |
| "learning_rate": 2.3766239334562452e-05, |
| "loss": 3.1942, |
| "step": 454500 |
| }, |
| { |
| "epoch": 19.179699026261435, |
| "grad_norm": 1.255203127861023, |
| "learning_rate": 2.3680487072846547e-05, |
| "loss": 3.1943, |
| "step": 455000 |
| }, |
| { |
| "epoch": 19.200775618597984, |
| "grad_norm": 1.2090405225753784, |
| "learning_rate": 2.3594734811130645e-05, |
| "loss": 3.1947, |
| "step": 455500 |
| }, |
| { |
| "epoch": 19.221852210934536, |
| "grad_norm": 1.2168551683425903, |
| "learning_rate": 2.3508982549414744e-05, |
| "loss": 3.1947, |
| "step": 456000 |
| }, |
| { |
| "epoch": 19.242928803271088, |
| "grad_norm": 1.2193043231964111, |
| "learning_rate": 2.342323028769884e-05, |
| "loss": 3.1986, |
| "step": 456500 |
| }, |
| { |
| "epoch": 19.264005395607636, |
| "grad_norm": 1.1891534328460693, |
| "learning_rate": 2.3337478025982937e-05, |
| "loss": 3.1943, |
| "step": 457000 |
| }, |
| { |
| "epoch": 19.28508198794419, |
| "grad_norm": 1.284074306488037, |
| "learning_rate": 2.3251725764267035e-05, |
| "loss": 3.1965, |
| "step": 457500 |
| }, |
| { |
| "epoch": 19.30615858028074, |
| "grad_norm": 1.2584196329116821, |
| "learning_rate": 2.316597350255113e-05, |
| "loss": 3.1985, |
| "step": 458000 |
| }, |
| { |
| "epoch": 19.327235172617293, |
| "grad_norm": 1.188183069229126, |
| "learning_rate": 2.308022124083523e-05, |
| "loss": 3.1959, |
| "step": 458500 |
| }, |
| { |
| "epoch": 19.34831176495384, |
| "grad_norm": 1.2135876417160034, |
| "learning_rate": 2.2994468979119323e-05, |
| "loss": 3.1976, |
| "step": 459000 |
| }, |
| { |
| "epoch": 19.369388357290394, |
| "grad_norm": 1.2653756141662598, |
| "learning_rate": 2.290871671740342e-05, |
| "loss": 3.1947, |
| "step": 459500 |
| }, |
| { |
| "epoch": 19.390464949626946, |
| "grad_norm": 1.2031307220458984, |
| "learning_rate": 2.282296445568752e-05, |
| "loss": 3.1962, |
| "step": 460000 |
| }, |
| { |
| "epoch": 19.390464949626946, |
| "eval_accuracy": 0.7547055658972496, |
| "eval_loss": 2.982649803161621, |
| "eval_runtime": 181.9405, |
| "eval_samples_per_second": 663.437, |
| "eval_steps_per_second": 6.914, |
| "step": 460000 |
| }, |
| { |
| "epoch": 19.411541541963494, |
| "grad_norm": 1.2776650190353394, |
| "learning_rate": 2.2737212193971615e-05, |
| "loss": 3.1946, |
| "step": 460500 |
| }, |
| { |
| "epoch": 19.432618134300046, |
| "grad_norm": 1.2217003107070923, |
| "learning_rate": 2.2651459932255713e-05, |
| "loss": 3.1982, |
| "step": 461000 |
| }, |
| { |
| "epoch": 19.4536947266366, |
| "grad_norm": 1.2361100912094116, |
| "learning_rate": 2.256570767053981e-05, |
| "loss": 3.1929, |
| "step": 461500 |
| }, |
| { |
| "epoch": 19.474771318973147, |
| "grad_norm": 1.2194640636444092, |
| "learning_rate": 2.2479955408823906e-05, |
| "loss": 3.1935, |
| "step": 462000 |
| }, |
| { |
| "epoch": 19.4958479113097, |
| "grad_norm": 1.2052329778671265, |
| "learning_rate": 2.2394203147108005e-05, |
| "loss": 3.1933, |
| "step": 462500 |
| }, |
| { |
| "epoch": 19.51692450364625, |
| "grad_norm": 1.1598479747772217, |
| "learning_rate": 2.2308450885392103e-05, |
| "loss": 3.1937, |
| "step": 463000 |
| }, |
| { |
| "epoch": 19.5380010959828, |
| "grad_norm": 1.2625000476837158, |
| "learning_rate": 2.22226986236762e-05, |
| "loss": 3.1917, |
| "step": 463500 |
| }, |
| { |
| "epoch": 19.559077688319352, |
| "grad_norm": 1.2167688608169556, |
| "learning_rate": 2.21369463619603e-05, |
| "loss": 3.1944, |
| "step": 464000 |
| }, |
| { |
| "epoch": 19.580154280655904, |
| "grad_norm": 1.312076449394226, |
| "learning_rate": 2.2051194100244394e-05, |
| "loss": 3.1929, |
| "step": 464500 |
| }, |
| { |
| "epoch": 19.601230872992456, |
| "grad_norm": 1.27461838722229, |
| "learning_rate": 2.1965441838528493e-05, |
| "loss": 3.1945, |
| "step": 465000 |
| }, |
| { |
| "epoch": 19.622307465329005, |
| "grad_norm": 1.2033019065856934, |
| "learning_rate": 2.187968957681259e-05, |
| "loss": 3.1931, |
| "step": 465500 |
| }, |
| { |
| "epoch": 19.643384057665557, |
| "grad_norm": 1.1653058528900146, |
| "learning_rate": 2.179393731509669e-05, |
| "loss": 3.1937, |
| "step": 466000 |
| }, |
| { |
| "epoch": 19.66446065000211, |
| "grad_norm": 1.2235221862792969, |
| "learning_rate": 2.1708185053380784e-05, |
| "loss": 3.1915, |
| "step": 466500 |
| }, |
| { |
| "epoch": 19.685537242338658, |
| "grad_norm": 1.2516303062438965, |
| "learning_rate": 2.1622432791664883e-05, |
| "loss": 3.195, |
| "step": 467000 |
| }, |
| { |
| "epoch": 19.70661383467521, |
| "grad_norm": 1.2479441165924072, |
| "learning_rate": 2.1536680529948978e-05, |
| "loss": 3.1937, |
| "step": 467500 |
| }, |
| { |
| "epoch": 19.727690427011762, |
| "grad_norm": 1.2611171007156372, |
| "learning_rate": 2.1450928268233076e-05, |
| "loss": 3.1922, |
| "step": 468000 |
| }, |
| { |
| "epoch": 19.74876701934831, |
| "grad_norm": 1.1851190328598022, |
| "learning_rate": 2.1365176006517174e-05, |
| "loss": 3.1927, |
| "step": 468500 |
| }, |
| { |
| "epoch": 19.769843611684863, |
| "grad_norm": 1.274307370185852, |
| "learning_rate": 2.127942374480127e-05, |
| "loss": 3.19, |
| "step": 469000 |
| }, |
| { |
| "epoch": 19.790920204021415, |
| "grad_norm": 1.2338905334472656, |
| "learning_rate": 2.1193671483085367e-05, |
| "loss": 3.1956, |
| "step": 469500 |
| }, |
| { |
| "epoch": 19.811996796357963, |
| "grad_norm": 1.2267173528671265, |
| "learning_rate": 2.1107919221369466e-05, |
| "loss": 3.1917, |
| "step": 470000 |
| }, |
| { |
| "epoch": 19.811996796357963, |
| "eval_accuracy": 0.7554824380411598, |
| "eval_loss": 2.9786159992218018, |
| "eval_runtime": 182.0746, |
| "eval_samples_per_second": 662.948, |
| "eval_steps_per_second": 6.909, |
| "step": 470000 |
| }, |
| { |
| "epoch": 19.833073388694515, |
| "grad_norm": 1.2134978771209717, |
| "learning_rate": 2.102216695965356e-05, |
| "loss": 3.1941, |
| "step": 470500 |
| }, |
| { |
| "epoch": 19.854149981031068, |
| "grad_norm": 1.2144368886947632, |
| "learning_rate": 2.093641469793766e-05, |
| "loss": 3.1913, |
| "step": 471000 |
| }, |
| { |
| "epoch": 19.87522657336762, |
| "grad_norm": 1.184815764427185, |
| "learning_rate": 2.0850662436221757e-05, |
| "loss": 3.1906, |
| "step": 471500 |
| }, |
| { |
| "epoch": 19.89630316570417, |
| "grad_norm": 1.220919132232666, |
| "learning_rate": 2.0764910174505852e-05, |
| "loss": 3.1915, |
| "step": 472000 |
| }, |
| { |
| "epoch": 19.91737975804072, |
| "grad_norm": 1.2284505367279053, |
| "learning_rate": 2.067915791278995e-05, |
| "loss": 3.1917, |
| "step": 472500 |
| }, |
| { |
| "epoch": 19.938456350377272, |
| "grad_norm": 1.2510801553726196, |
| "learning_rate": 2.059340565107405e-05, |
| "loss": 3.1862, |
| "step": 473000 |
| }, |
| { |
| "epoch": 19.95953294271382, |
| "grad_norm": 1.2268187999725342, |
| "learning_rate": 2.0507653389358144e-05, |
| "loss": 3.1902, |
| "step": 473500 |
| }, |
| { |
| "epoch": 19.980609535050373, |
| "grad_norm": 1.2784428596496582, |
| "learning_rate": 2.0421901127642242e-05, |
| "loss": 3.1874, |
| "step": 474000 |
| }, |
| { |
| "epoch": 20.001686127386925, |
| "grad_norm": 1.2232112884521484, |
| "learning_rate": 2.033614886592634e-05, |
| "loss": 3.1961, |
| "step": 474500 |
| }, |
| { |
| "epoch": 20.022762719723474, |
| "grad_norm": 1.2236319780349731, |
| "learning_rate": 2.0250396604210435e-05, |
| "loss": 3.1887, |
| "step": 475000 |
| }, |
| { |
| "epoch": 20.043839312060026, |
| "grad_norm": 1.2405495643615723, |
| "learning_rate": 2.0164644342494534e-05, |
| "loss": 3.1887, |
| "step": 475500 |
| }, |
| { |
| "epoch": 20.064915904396578, |
| "grad_norm": 1.2011206150054932, |
| "learning_rate": 2.007889208077863e-05, |
| "loss": 3.181, |
| "step": 476000 |
| }, |
| { |
| "epoch": 20.085992496733127, |
| "grad_norm": 1.2511368989944458, |
| "learning_rate": 1.9993139819062727e-05, |
| "loss": 3.1858, |
| "step": 476500 |
| }, |
| { |
| "epoch": 20.10706908906968, |
| "grad_norm": 1.2314043045043945, |
| "learning_rate": 1.9907387557346825e-05, |
| "loss": 3.1868, |
| "step": 477000 |
| }, |
| { |
| "epoch": 20.12814568140623, |
| "grad_norm": 1.2037440538406372, |
| "learning_rate": 1.9821635295630923e-05, |
| "loss": 3.1852, |
| "step": 477500 |
| }, |
| { |
| "epoch": 20.14922227374278, |
| "grad_norm": 1.2563297748565674, |
| "learning_rate": 1.9735883033915022e-05, |
| "loss": 3.1852, |
| "step": 478000 |
| }, |
| { |
| "epoch": 20.17029886607933, |
| "grad_norm": 1.2514755725860596, |
| "learning_rate": 1.965013077219912e-05, |
| "loss": 3.1875, |
| "step": 478500 |
| }, |
| { |
| "epoch": 20.191375458415884, |
| "grad_norm": 1.245760202407837, |
| "learning_rate": 1.9564378510483215e-05, |
| "loss": 3.185, |
| "step": 479000 |
| }, |
| { |
| "epoch": 20.212452050752436, |
| "grad_norm": 1.1998268365859985, |
| "learning_rate": 1.9478626248767313e-05, |
| "loss": 3.1894, |
| "step": 479500 |
| }, |
| { |
| "epoch": 20.233528643088984, |
| "grad_norm": 1.2781301736831665, |
| "learning_rate": 1.939287398705141e-05, |
| "loss": 3.1854, |
| "step": 480000 |
| }, |
| { |
| "epoch": 20.233528643088984, |
| "eval_accuracy": 0.7560993630389371, |
| "eval_loss": 2.975541353225708, |
| "eval_runtime": 181.8413, |
| "eval_samples_per_second": 663.799, |
| "eval_steps_per_second": 6.918, |
| "step": 480000 |
| }, |
| { |
| "epoch": 20.254605235425537, |
| "grad_norm": 1.17195725440979, |
| "learning_rate": 1.9307121725335506e-05, |
| "loss": 3.1828, |
| "step": 480500 |
| }, |
| { |
| "epoch": 20.27568182776209, |
| "grad_norm": 1.3203258514404297, |
| "learning_rate": 1.9221369463619605e-05, |
| "loss": 3.1813, |
| "step": 481000 |
| }, |
| { |
| "epoch": 20.296758420098637, |
| "grad_norm": 1.23111891746521, |
| "learning_rate": 1.9135617201903703e-05, |
| "loss": 3.1824, |
| "step": 481500 |
| }, |
| { |
| "epoch": 20.31783501243519, |
| "grad_norm": 1.2626653909683228, |
| "learning_rate": 1.9049864940187798e-05, |
| "loss": 3.1833, |
| "step": 482000 |
| }, |
| { |
| "epoch": 20.33891160477174, |
| "grad_norm": 1.2856457233428955, |
| "learning_rate": 1.8964112678471896e-05, |
| "loss": 3.186, |
| "step": 482500 |
| }, |
| { |
| "epoch": 20.35998819710829, |
| "grad_norm": 1.2776159048080444, |
| "learning_rate": 1.8878360416755995e-05, |
| "loss": 3.1867, |
| "step": 483000 |
| }, |
| { |
| "epoch": 20.381064789444842, |
| "grad_norm": 1.2357022762298584, |
| "learning_rate": 1.879260815504009e-05, |
| "loss": 3.1837, |
| "step": 483500 |
| }, |
| { |
| "epoch": 20.402141381781394, |
| "grad_norm": 1.2061140537261963, |
| "learning_rate": 1.8706855893324188e-05, |
| "loss": 3.1821, |
| "step": 484000 |
| }, |
| { |
| "epoch": 20.423217974117943, |
| "grad_norm": 1.228149652481079, |
| "learning_rate": 1.8621103631608283e-05, |
| "loss": 3.1851, |
| "step": 484500 |
| }, |
| { |
| "epoch": 20.444294566454495, |
| "grad_norm": 1.2854108810424805, |
| "learning_rate": 1.853535136989238e-05, |
| "loss": 3.1822, |
| "step": 485000 |
| }, |
| { |
| "epoch": 20.465371158791047, |
| "grad_norm": 1.2384010553359985, |
| "learning_rate": 1.844959910817648e-05, |
| "loss": 3.1819, |
| "step": 485500 |
| }, |
| { |
| "epoch": 20.4864477511276, |
| "grad_norm": 1.2463432550430298, |
| "learning_rate": 1.8363846846460574e-05, |
| "loss": 3.1854, |
| "step": 486000 |
| }, |
| { |
| "epoch": 20.507524343464148, |
| "grad_norm": 1.2451106309890747, |
| "learning_rate": 1.8278094584744673e-05, |
| "loss": 3.1798, |
| "step": 486500 |
| }, |
| { |
| "epoch": 20.5286009358007, |
| "grad_norm": 1.2418768405914307, |
| "learning_rate": 1.819234232302877e-05, |
| "loss": 3.184, |
| "step": 487000 |
| }, |
| { |
| "epoch": 20.549677528137252, |
| "grad_norm": 1.2279064655303955, |
| "learning_rate": 1.8106590061312866e-05, |
| "loss": 3.1867, |
| "step": 487500 |
| }, |
| { |
| "epoch": 20.5707541204738, |
| "grad_norm": 1.2425814867019653, |
| "learning_rate": 1.8020837799596964e-05, |
| "loss": 3.1849, |
| "step": 488000 |
| }, |
| { |
| "epoch": 20.591830712810353, |
| "grad_norm": 1.2582523822784424, |
| "learning_rate": 1.7935085537881062e-05, |
| "loss": 3.1835, |
| "step": 488500 |
| }, |
| { |
| "epoch": 20.612907305146905, |
| "grad_norm": 1.3201396465301514, |
| "learning_rate": 1.7849333276165157e-05, |
| "loss": 3.1832, |
| "step": 489000 |
| }, |
| { |
| "epoch": 20.633983897483454, |
| "grad_norm": 1.2965776920318604, |
| "learning_rate": 1.7763581014449256e-05, |
| "loss": 3.182, |
| "step": 489500 |
| }, |
| { |
| "epoch": 20.655060489820006, |
| "grad_norm": 1.2087916135787964, |
| "learning_rate": 1.7677828752733354e-05, |
| "loss": 3.18, |
| "step": 490000 |
| }, |
| { |
| "epoch": 20.655060489820006, |
| "eval_accuracy": 0.7566523515404342, |
| "eval_loss": 2.9710960388183594, |
| "eval_runtime": 182.4164, |
| "eval_samples_per_second": 661.706, |
| "eval_steps_per_second": 6.896, |
| "step": 490000 |
| }, |
| { |
| "epoch": 20.676137082156558, |
| "grad_norm": 1.265891432762146, |
| "learning_rate": 1.7592076491017452e-05, |
| "loss": 3.1829, |
| "step": 490500 |
| }, |
| { |
| "epoch": 20.697213674493106, |
| "grad_norm": 1.2662580013275146, |
| "learning_rate": 1.7506324229301547e-05, |
| "loss": 3.1838, |
| "step": 491000 |
| }, |
| { |
| "epoch": 20.71829026682966, |
| "grad_norm": 1.2259025573730469, |
| "learning_rate": 1.7420571967585646e-05, |
| "loss": 3.1848, |
| "step": 491500 |
| }, |
| { |
| "epoch": 20.73936685916621, |
| "grad_norm": 1.2641977071762085, |
| "learning_rate": 1.7334819705869744e-05, |
| "loss": 3.1815, |
| "step": 492000 |
| }, |
| { |
| "epoch": 20.760443451502763, |
| "grad_norm": 1.238683819770813, |
| "learning_rate": 1.7249067444153842e-05, |
| "loss": 3.1846, |
| "step": 492500 |
| }, |
| { |
| "epoch": 20.78152004383931, |
| "grad_norm": 1.226242184638977, |
| "learning_rate": 1.7163315182437937e-05, |
| "loss": 3.1788, |
| "step": 493000 |
| }, |
| { |
| "epoch": 20.802596636175863, |
| "grad_norm": 1.2646241188049316, |
| "learning_rate": 1.7077562920722035e-05, |
| "loss": 3.179, |
| "step": 493500 |
| }, |
| { |
| "epoch": 20.823673228512416, |
| "grad_norm": 1.236530065536499, |
| "learning_rate": 1.6991810659006134e-05, |
| "loss": 3.1814, |
| "step": 494000 |
| }, |
| { |
| "epoch": 20.844749820848964, |
| "grad_norm": 1.2542351484298706, |
| "learning_rate": 1.690605839729023e-05, |
| "loss": 3.1814, |
| "step": 494500 |
| }, |
| { |
| "epoch": 20.865826413185516, |
| "grad_norm": 1.1883656978607178, |
| "learning_rate": 1.6820306135574327e-05, |
| "loss": 3.1819, |
| "step": 495000 |
| }, |
| { |
| "epoch": 20.88690300552207, |
| "grad_norm": 1.2520887851715088, |
| "learning_rate": 1.6734553873858425e-05, |
| "loss": 3.1827, |
| "step": 495500 |
| }, |
| { |
| "epoch": 20.907979597858617, |
| "grad_norm": 1.2592312097549438, |
| "learning_rate": 1.664880161214252e-05, |
| "loss": 3.1811, |
| "step": 496000 |
| }, |
| { |
| "epoch": 20.92905619019517, |
| "grad_norm": 1.2282456159591675, |
| "learning_rate": 1.656304935042662e-05, |
| "loss": 3.1827, |
| "step": 496500 |
| }, |
| { |
| "epoch": 20.95013278253172, |
| "grad_norm": 1.2040045261383057, |
| "learning_rate": 1.6477297088710717e-05, |
| "loss": 3.1792, |
| "step": 497000 |
| }, |
| { |
| "epoch": 20.97120937486827, |
| "grad_norm": 1.2309666872024536, |
| "learning_rate": 1.639154482699481e-05, |
| "loss": 3.1806, |
| "step": 497500 |
| }, |
| { |
| "epoch": 20.992285967204822, |
| "grad_norm": 1.22409188747406, |
| "learning_rate": 1.630579256527891e-05, |
| "loss": 3.1786, |
| "step": 498000 |
| }, |
| { |
| "epoch": 21.013362559541374, |
| "grad_norm": 1.2257660627365112, |
| "learning_rate": 1.6220040303563008e-05, |
| "loss": 3.1782, |
| "step": 498500 |
| }, |
| { |
| "epoch": 21.034439151877926, |
| "grad_norm": 1.2467764616012573, |
| "learning_rate": 1.6134288041847103e-05, |
| "loss": 3.1808, |
| "step": 499000 |
| }, |
| { |
| "epoch": 21.055515744214475, |
| "grad_norm": 1.2346512079238892, |
| "learning_rate": 1.60485357801312e-05, |
| "loss": 3.1764, |
| "step": 499500 |
| }, |
| { |
| "epoch": 21.076592336551027, |
| "grad_norm": 1.2651658058166504, |
| "learning_rate": 1.59627835184153e-05, |
| "loss": 3.1744, |
| "step": 500000 |
| }, |
| { |
| "epoch": 21.076592336551027, |
| "eval_accuracy": 0.7573346645425043, |
| "eval_loss": 2.9676096439361572, |
| "eval_runtime": 179.7949, |
| "eval_samples_per_second": 671.354, |
| "eval_steps_per_second": 6.997, |
| "step": 500000 |
| }, |
| { |
| "epoch": 21.09766892888758, |
| "grad_norm": 1.235007405281067, |
| "learning_rate": 1.5877031256699395e-05, |
| "loss": 3.1763, |
| "step": 500500 |
| }, |
| { |
| "epoch": 21.118745521224128, |
| "grad_norm": 1.3721537590026855, |
| "learning_rate": 1.5791278994983493e-05, |
| "loss": 3.1785, |
| "step": 501000 |
| }, |
| { |
| "epoch": 21.13982211356068, |
| "grad_norm": 1.1944681406021118, |
| "learning_rate": 1.570552673326759e-05, |
| "loss": 3.1743, |
| "step": 501500 |
| }, |
| { |
| "epoch": 21.16089870589723, |
| "grad_norm": 1.2210240364074707, |
| "learning_rate": 1.5619774471551686e-05, |
| "loss": 3.1786, |
| "step": 502000 |
| }, |
| { |
| "epoch": 21.18197529823378, |
| "grad_norm": 1.1890056133270264, |
| "learning_rate": 1.5534022209835785e-05, |
| "loss": 3.1771, |
| "step": 502500 |
| }, |
| { |
| "epoch": 21.203051890570332, |
| "grad_norm": 1.171532154083252, |
| "learning_rate": 1.544826994811988e-05, |
| "loss": 3.1774, |
| "step": 503000 |
| }, |
| { |
| "epoch": 21.224128482906885, |
| "grad_norm": 1.2562947273254395, |
| "learning_rate": 1.5362517686403978e-05, |
| "loss": 3.1773, |
| "step": 503500 |
| }, |
| { |
| "epoch": 21.245205075243433, |
| "grad_norm": 1.2311848402023315, |
| "learning_rate": 1.5276765424688076e-05, |
| "loss": 3.1762, |
| "step": 504000 |
| }, |
| { |
| "epoch": 21.266281667579985, |
| "grad_norm": 1.2791424989700317, |
| "learning_rate": 1.5191013162972173e-05, |
| "loss": 3.1757, |
| "step": 504500 |
| }, |
| { |
| "epoch": 21.287358259916537, |
| "grad_norm": 1.251415729522705, |
| "learning_rate": 1.5105260901256271e-05, |
| "loss": 3.1761, |
| "step": 505000 |
| }, |
| { |
| "epoch": 21.308434852253086, |
| "grad_norm": 1.267265796661377, |
| "learning_rate": 1.501950863954037e-05, |
| "loss": 3.1762, |
| "step": 505500 |
| }, |
| { |
| "epoch": 21.329511444589638, |
| "grad_norm": 1.2331628799438477, |
| "learning_rate": 1.4933756377824464e-05, |
| "loss": 3.1737, |
| "step": 506000 |
| }, |
| { |
| "epoch": 21.35058803692619, |
| "grad_norm": 1.2210456132888794, |
| "learning_rate": 1.4848004116108563e-05, |
| "loss": 3.1701, |
| "step": 506500 |
| }, |
| { |
| "epoch": 21.371664629262742, |
| "grad_norm": 1.187347412109375, |
| "learning_rate": 1.4762251854392661e-05, |
| "loss": 3.1774, |
| "step": 507000 |
| }, |
| { |
| "epoch": 21.39274122159929, |
| "grad_norm": 1.2321112155914307, |
| "learning_rate": 1.4676499592676758e-05, |
| "loss": 3.1763, |
| "step": 507500 |
| }, |
| { |
| "epoch": 21.413817813935843, |
| "grad_norm": 1.2151174545288086, |
| "learning_rate": 1.4590747330960856e-05, |
| "loss": 3.1741, |
| "step": 508000 |
| }, |
| { |
| "epoch": 21.434894406272395, |
| "grad_norm": 1.2247110605239868, |
| "learning_rate": 1.4504995069244952e-05, |
| "loss": 3.1703, |
| "step": 508500 |
| }, |
| { |
| "epoch": 21.455970998608944, |
| "grad_norm": 1.166321039199829, |
| "learning_rate": 1.4419242807529049e-05, |
| "loss": 3.1749, |
| "step": 509000 |
| }, |
| { |
| "epoch": 21.477047590945496, |
| "grad_norm": 1.189142107963562, |
| "learning_rate": 1.4333490545813147e-05, |
| "loss": 3.175, |
| "step": 509500 |
| }, |
| { |
| "epoch": 21.498124183282048, |
| "grad_norm": 1.282515525817871, |
| "learning_rate": 1.4247738284097246e-05, |
| "loss": 3.1714, |
| "step": 510000 |
| }, |
| { |
| "epoch": 21.498124183282048, |
| "eval_accuracy": 0.7580059163544242, |
| "eval_loss": 2.963520050048828, |
| "eval_runtime": 182.0438, |
| "eval_samples_per_second": 663.06, |
| "eval_steps_per_second": 6.91, |
| "step": 510000 |
| }, |
| { |
| "epoch": 21.519200775618597, |
| "grad_norm": 1.2143868207931519, |
| "learning_rate": 1.416198602238134e-05, |
| "loss": 3.1747, |
| "step": 510500 |
| }, |
| { |
| "epoch": 21.54027736795515, |
| "grad_norm": 1.1803776025772095, |
| "learning_rate": 1.4076233760665439e-05, |
| "loss": 3.1731, |
| "step": 511000 |
| }, |
| { |
| "epoch": 21.5613539602917, |
| "grad_norm": 1.272083044052124, |
| "learning_rate": 1.3990481498949534e-05, |
| "loss": 3.1709, |
| "step": 511500 |
| }, |
| { |
| "epoch": 21.58243055262825, |
| "grad_norm": 1.2092067003250122, |
| "learning_rate": 1.3904729237233632e-05, |
| "loss": 3.1749, |
| "step": 512000 |
| }, |
| { |
| "epoch": 21.6035071449648, |
| "grad_norm": 1.2079823017120361, |
| "learning_rate": 1.381897697551773e-05, |
| "loss": 3.1716, |
| "step": 512500 |
| }, |
| { |
| "epoch": 21.624583737301354, |
| "grad_norm": 1.2442198991775513, |
| "learning_rate": 1.3733224713801825e-05, |
| "loss": 3.1758, |
| "step": 513000 |
| }, |
| { |
| "epoch": 21.645660329637906, |
| "grad_norm": 1.2682937383651733, |
| "learning_rate": 1.3647472452085924e-05, |
| "loss": 3.1794, |
| "step": 513500 |
| }, |
| { |
| "epoch": 21.666736921974454, |
| "grad_norm": 1.269411563873291, |
| "learning_rate": 1.3561720190370022e-05, |
| "loss": 3.1731, |
| "step": 514000 |
| }, |
| { |
| "epoch": 21.687813514311006, |
| "grad_norm": 1.2496118545532227, |
| "learning_rate": 1.3475967928654119e-05, |
| "loss": 3.1703, |
| "step": 514500 |
| }, |
| { |
| "epoch": 21.70889010664756, |
| "grad_norm": 1.259488821029663, |
| "learning_rate": 1.3390215666938217e-05, |
| "loss": 3.1742, |
| "step": 515000 |
| }, |
| { |
| "epoch": 21.729966698984107, |
| "grad_norm": 1.2047630548477173, |
| "learning_rate": 1.3304463405222315e-05, |
| "loss": 3.1742, |
| "step": 515500 |
| }, |
| { |
| "epoch": 21.75104329132066, |
| "grad_norm": 1.3011350631713867, |
| "learning_rate": 1.321871114350641e-05, |
| "loss": 3.1748, |
| "step": 516000 |
| }, |
| { |
| "epoch": 21.77211988365721, |
| "grad_norm": 1.2550950050354004, |
| "learning_rate": 1.3132958881790508e-05, |
| "loss": 3.1708, |
| "step": 516500 |
| }, |
| { |
| "epoch": 21.79319647599376, |
| "grad_norm": 1.2112140655517578, |
| "learning_rate": 1.3047206620074607e-05, |
| "loss": 3.1698, |
| "step": 517000 |
| }, |
| { |
| "epoch": 21.814273068330312, |
| "grad_norm": 1.29332435131073, |
| "learning_rate": 1.2961454358358702e-05, |
| "loss": 3.172, |
| "step": 517500 |
| }, |
| { |
| "epoch": 21.835349660666864, |
| "grad_norm": 1.3130245208740234, |
| "learning_rate": 1.28757020966428e-05, |
| "loss": 3.1679, |
| "step": 518000 |
| }, |
| { |
| "epoch": 21.856426253003413, |
| "grad_norm": 1.2165025472640991, |
| "learning_rate": 1.2789949834926898e-05, |
| "loss": 3.1681, |
| "step": 518500 |
| }, |
| { |
| "epoch": 21.877502845339965, |
| "grad_norm": 1.2566921710968018, |
| "learning_rate": 1.2704197573210993e-05, |
| "loss": 3.1751, |
| "step": 519000 |
| }, |
| { |
| "epoch": 21.898579437676517, |
| "grad_norm": 1.2312777042388916, |
| "learning_rate": 1.2618445311495092e-05, |
| "loss": 3.1688, |
| "step": 519500 |
| }, |
| { |
| "epoch": 21.91965603001307, |
| "grad_norm": 1.2554585933685303, |
| "learning_rate": 1.2532693049779186e-05, |
| "loss": 3.1664, |
| "step": 520000 |
| }, |
| { |
| "epoch": 21.91965603001307, |
| "eval_accuracy": 0.7583284001726168, |
| "eval_loss": 2.961617946624756, |
| "eval_runtime": 182.0507, |
| "eval_samples_per_second": 663.035, |
| "eval_steps_per_second": 6.91, |
| "step": 520000 |
| }, |
| { |
| "epoch": 21.940732622349618, |
| "grad_norm": 1.3006466627120972, |
| "learning_rate": 1.2446940788063285e-05, |
| "loss": 3.1721, |
| "step": 520500 |
| }, |
| { |
| "epoch": 21.96180921468617, |
| "grad_norm": 1.2451550960540771, |
| "learning_rate": 1.2361188526347383e-05, |
| "loss": 3.1708, |
| "step": 521000 |
| }, |
| { |
| "epoch": 21.982885807022722, |
| "grad_norm": 1.2090649604797363, |
| "learning_rate": 1.227543626463148e-05, |
| "loss": 3.1712, |
| "step": 521500 |
| }, |
| { |
| "epoch": 22.00396239935927, |
| "grad_norm": 1.3153047561645508, |
| "learning_rate": 1.2189684002915578e-05, |
| "loss": 3.1681, |
| "step": 522000 |
| }, |
| { |
| "epoch": 22.025038991695823, |
| "grad_norm": 1.3064744472503662, |
| "learning_rate": 1.2103931741199675e-05, |
| "loss": 3.1655, |
| "step": 522500 |
| }, |
| { |
| "epoch": 22.046115584032375, |
| "grad_norm": 1.2643072605133057, |
| "learning_rate": 1.2018179479483773e-05, |
| "loss": 3.1675, |
| "step": 523000 |
| }, |
| { |
| "epoch": 22.067192176368923, |
| "grad_norm": 1.2299425601959229, |
| "learning_rate": 1.193242721776787e-05, |
| "loss": 3.1688, |
| "step": 523500 |
| }, |
| { |
| "epoch": 22.088268768705476, |
| "grad_norm": 1.2600406408309937, |
| "learning_rate": 1.1846674956051966e-05, |
| "loss": 3.1702, |
| "step": 524000 |
| }, |
| { |
| "epoch": 22.109345361042028, |
| "grad_norm": 1.2509068250656128, |
| "learning_rate": 1.1760922694336064e-05, |
| "loss": 3.1658, |
| "step": 524500 |
| }, |
| { |
| "epoch": 22.130421953378576, |
| "grad_norm": 1.2505934238433838, |
| "learning_rate": 1.1675170432620161e-05, |
| "loss": 3.1705, |
| "step": 525000 |
| }, |
| { |
| "epoch": 22.15149854571513, |
| "grad_norm": 1.216555118560791, |
| "learning_rate": 1.1589418170904258e-05, |
| "loss": 3.1644, |
| "step": 525500 |
| }, |
| { |
| "epoch": 22.17257513805168, |
| "grad_norm": 1.2362140417099, |
| "learning_rate": 1.1503665909188354e-05, |
| "loss": 3.1656, |
| "step": 526000 |
| }, |
| { |
| "epoch": 22.193651730388233, |
| "grad_norm": 1.2833468914031982, |
| "learning_rate": 1.1417913647472453e-05, |
| "loss": 3.1663, |
| "step": 526500 |
| }, |
| { |
| "epoch": 22.21472832272478, |
| "grad_norm": 1.2725893259048462, |
| "learning_rate": 1.133216138575655e-05, |
| "loss": 3.168, |
| "step": 527000 |
| }, |
| { |
| "epoch": 22.235804915061333, |
| "grad_norm": 1.2721750736236572, |
| "learning_rate": 1.1246409124040646e-05, |
| "loss": 3.1639, |
| "step": 527500 |
| }, |
| { |
| "epoch": 22.256881507397885, |
| "grad_norm": 1.2804903984069824, |
| "learning_rate": 1.1160656862324744e-05, |
| "loss": 3.1681, |
| "step": 528000 |
| }, |
| { |
| "epoch": 22.277958099734434, |
| "grad_norm": 1.248216152191162, |
| "learning_rate": 1.1074904600608842e-05, |
| "loss": 3.1667, |
| "step": 528500 |
| }, |
| { |
| "epoch": 22.299034692070986, |
| "grad_norm": 1.2912533283233643, |
| "learning_rate": 1.0989152338892939e-05, |
| "loss": 3.1637, |
| "step": 529000 |
| }, |
| { |
| "epoch": 22.320111284407538, |
| "grad_norm": 1.2848520278930664, |
| "learning_rate": 1.0903400077177037e-05, |
| "loss": 3.1677, |
| "step": 529500 |
| }, |
| { |
| "epoch": 22.341187876744087, |
| "grad_norm": 1.2185943126678467, |
| "learning_rate": 1.0817647815461134e-05, |
| "loss": 3.1674, |
| "step": 530000 |
| }, |
| { |
| "epoch": 22.341187876744087, |
| "eval_accuracy": 0.7587943656280662, |
| "eval_loss": 2.9592065811157227, |
| "eval_runtime": 181.6591, |
| "eval_samples_per_second": 664.464, |
| "eval_steps_per_second": 6.925, |
| "step": 530000 |
| }, |
| { |
| "epoch": 22.36226446908064, |
| "grad_norm": 1.2643219232559204, |
| "learning_rate": 1.073189555374523e-05, |
| "loss": 3.1677, |
| "step": 530500 |
| }, |
| { |
| "epoch": 22.38334106141719, |
| "grad_norm": 1.3079557418823242, |
| "learning_rate": 1.0646143292029327e-05, |
| "loss": 3.1693, |
| "step": 531000 |
| }, |
| { |
| "epoch": 22.40441765375374, |
| "grad_norm": 1.2448381185531616, |
| "learning_rate": 1.0560391030313425e-05, |
| "loss": 3.1633, |
| "step": 531500 |
| }, |
| { |
| "epoch": 22.42549424609029, |
| "grad_norm": 1.2501459121704102, |
| "learning_rate": 1.0474638768597522e-05, |
| "loss": 3.1621, |
| "step": 532000 |
| }, |
| { |
| "epoch": 22.446570838426844, |
| "grad_norm": 1.2736622095108032, |
| "learning_rate": 1.0388886506881619e-05, |
| "loss": 3.1652, |
| "step": 532500 |
| }, |
| { |
| "epoch": 22.467647430763392, |
| "grad_norm": 1.2683038711547852, |
| "learning_rate": 1.0303134245165717e-05, |
| "loss": 3.1672, |
| "step": 533000 |
| }, |
| { |
| "epoch": 22.488724023099945, |
| "grad_norm": 1.2822680473327637, |
| "learning_rate": 1.0217381983449814e-05, |
| "loss": 3.1652, |
| "step": 533500 |
| }, |
| { |
| "epoch": 22.509800615436497, |
| "grad_norm": 1.2086163759231567, |
| "learning_rate": 1.013162972173391e-05, |
| "loss": 3.1651, |
| "step": 534000 |
| }, |
| { |
| "epoch": 22.53087720777305, |
| "grad_norm": 1.2650340795516968, |
| "learning_rate": 1.0045877460018009e-05, |
| "loss": 3.1679, |
| "step": 534500 |
| }, |
| { |
| "epoch": 22.551953800109597, |
| "grad_norm": 1.2698005437850952, |
| "learning_rate": 9.960125198302105e-06, |
| "loss": 3.1698, |
| "step": 535000 |
| }, |
| { |
| "epoch": 22.57303039244615, |
| "grad_norm": 1.222213864326477, |
| "learning_rate": 9.874372936586203e-06, |
| "loss": 3.1642, |
| "step": 535500 |
| }, |
| { |
| "epoch": 22.5941069847827, |
| "grad_norm": 1.2549235820770264, |
| "learning_rate": 9.7886206748703e-06, |
| "loss": 3.1639, |
| "step": 536000 |
| }, |
| { |
| "epoch": 22.61518357711925, |
| "grad_norm": 1.2828950881958008, |
| "learning_rate": 9.702868413154398e-06, |
| "loss": 3.1639, |
| "step": 536500 |
| }, |
| { |
| "epoch": 22.636260169455802, |
| "grad_norm": 1.2685843706130981, |
| "learning_rate": 9.617116151438495e-06, |
| "loss": 3.1631, |
| "step": 537000 |
| }, |
| { |
| "epoch": 22.657336761792354, |
| "grad_norm": 1.2327650785446167, |
| "learning_rate": 9.531363889722592e-06, |
| "loss": 3.165, |
| "step": 537500 |
| }, |
| { |
| "epoch": 22.678413354128903, |
| "grad_norm": 1.2177008390426636, |
| "learning_rate": 9.44561162800669e-06, |
| "loss": 3.1637, |
| "step": 538000 |
| }, |
| { |
| "epoch": 22.699489946465455, |
| "grad_norm": 1.229316234588623, |
| "learning_rate": 9.359859366290787e-06, |
| "loss": 3.1657, |
| "step": 538500 |
| }, |
| { |
| "epoch": 22.720566538802007, |
| "grad_norm": 1.183287262916565, |
| "learning_rate": 9.274107104574883e-06, |
| "loss": 3.166, |
| "step": 539000 |
| }, |
| { |
| "epoch": 22.741643131138556, |
| "grad_norm": 1.2221888303756714, |
| "learning_rate": 9.18835484285898e-06, |
| "loss": 3.1609, |
| "step": 539500 |
| }, |
| { |
| "epoch": 22.762719723475108, |
| "grad_norm": 1.247989296913147, |
| "learning_rate": 9.102602581143078e-06, |
| "loss": 3.1614, |
| "step": 540000 |
| }, |
| { |
| "epoch": 22.762719723475108, |
| "eval_accuracy": 0.7590821489932136, |
| "eval_loss": 2.957890510559082, |
| "eval_runtime": 183.0437, |
| "eval_samples_per_second": 659.438, |
| "eval_steps_per_second": 6.873, |
| "step": 540000 |
| }, |
| { |
| "epoch": 22.78379631581166, |
| "grad_norm": 1.2804683446884155, |
| "learning_rate": 9.016850319427175e-06, |
| "loss": 3.1652, |
| "step": 540500 |
| }, |
| { |
| "epoch": 22.804872908148212, |
| "grad_norm": 1.263993740081787, |
| "learning_rate": 8.931098057711271e-06, |
| "loss": 3.1649, |
| "step": 541000 |
| }, |
| { |
| "epoch": 22.82594950048476, |
| "grad_norm": 1.2425966262817383, |
| "learning_rate": 8.84534579599537e-06, |
| "loss": 3.1637, |
| "step": 541500 |
| }, |
| { |
| "epoch": 22.847026092821313, |
| "grad_norm": 1.2995390892028809, |
| "learning_rate": 8.759593534279468e-06, |
| "loss": 3.1642, |
| "step": 542000 |
| }, |
| { |
| "epoch": 22.868102685157865, |
| "grad_norm": 1.1968458890914917, |
| "learning_rate": 8.673841272563565e-06, |
| "loss": 3.1638, |
| "step": 542500 |
| }, |
| { |
| "epoch": 22.889179277494414, |
| "grad_norm": 1.2422608137130737, |
| "learning_rate": 8.588089010847663e-06, |
| "loss": 3.165, |
| "step": 543000 |
| }, |
| { |
| "epoch": 22.910255869830966, |
| "grad_norm": 1.2811158895492554, |
| "learning_rate": 8.50233674913176e-06, |
| "loss": 3.1631, |
| "step": 543500 |
| }, |
| { |
| "epoch": 22.931332462167518, |
| "grad_norm": 1.2808364629745483, |
| "learning_rate": 8.416584487415856e-06, |
| "loss": 3.1591, |
| "step": 544000 |
| }, |
| { |
| "epoch": 22.952409054504066, |
| "grad_norm": 1.3183255195617676, |
| "learning_rate": 8.330832225699953e-06, |
| "loss": 3.1586, |
| "step": 544500 |
| }, |
| { |
| "epoch": 22.97348564684062, |
| "grad_norm": 1.2929859161376953, |
| "learning_rate": 8.245079963984051e-06, |
| "loss": 3.1587, |
| "step": 545000 |
| }, |
| { |
| "epoch": 22.99456223917717, |
| "grad_norm": 1.2606472969055176, |
| "learning_rate": 8.159327702268148e-06, |
| "loss": 3.1636, |
| "step": 545500 |
| }, |
| { |
| "epoch": 23.01563883151372, |
| "grad_norm": 1.249458909034729, |
| "learning_rate": 8.073575440552244e-06, |
| "loss": 3.1612, |
| "step": 546000 |
| }, |
| { |
| "epoch": 23.03671542385027, |
| "grad_norm": 1.2890489101409912, |
| "learning_rate": 7.987823178836343e-06, |
| "loss": 3.1607, |
| "step": 546500 |
| }, |
| { |
| "epoch": 23.057792016186824, |
| "grad_norm": 1.2749804258346558, |
| "learning_rate": 7.90207091712044e-06, |
| "loss": 3.16, |
| "step": 547000 |
| }, |
| { |
| "epoch": 23.078868608523376, |
| "grad_norm": 1.2637490034103394, |
| "learning_rate": 7.816318655404536e-06, |
| "loss": 3.1573, |
| "step": 547500 |
| }, |
| { |
| "epoch": 23.099945200859924, |
| "grad_norm": 1.2606751918792725, |
| "learning_rate": 7.730566393688634e-06, |
| "loss": 3.1606, |
| "step": 548000 |
| }, |
| { |
| "epoch": 23.121021793196476, |
| "grad_norm": 1.2226364612579346, |
| "learning_rate": 7.64481413197273e-06, |
| "loss": 3.1583, |
| "step": 548500 |
| }, |
| { |
| "epoch": 23.14209838553303, |
| "grad_norm": 1.2156528234481812, |
| "learning_rate": 7.559061870256828e-06, |
| "loss": 3.1609, |
| "step": 549000 |
| }, |
| { |
| "epoch": 23.163174977869577, |
| "grad_norm": 1.243517518043518, |
| "learning_rate": 7.473309608540925e-06, |
| "loss": 3.1582, |
| "step": 549500 |
| }, |
| { |
| "epoch": 23.18425157020613, |
| "grad_norm": 1.292385220527649, |
| "learning_rate": 7.387557346825023e-06, |
| "loss": 3.1616, |
| "step": 550000 |
| }, |
| { |
| "epoch": 23.18425157020613, |
| "eval_accuracy": 0.7597256515171029, |
| "eval_loss": 2.953620672225952, |
| "eval_runtime": 180.2754, |
| "eval_samples_per_second": 669.564, |
| "eval_steps_per_second": 6.978, |
| "step": 550000 |
| }, |
| { |
| "epoch": 23.20532816254268, |
| "grad_norm": 1.309609293937683, |
| "learning_rate": 7.3018050851091205e-06, |
| "loss": 3.1569, |
| "step": 550500 |
| }, |
| { |
| "epoch": 23.22640475487923, |
| "grad_norm": 1.2534027099609375, |
| "learning_rate": 7.216052823393217e-06, |
| "loss": 3.1601, |
| "step": 551000 |
| }, |
| { |
| "epoch": 23.247481347215782, |
| "grad_norm": 1.2721811532974243, |
| "learning_rate": 7.1303005616773155e-06, |
| "loss": 3.1591, |
| "step": 551500 |
| }, |
| { |
| "epoch": 23.268557939552334, |
| "grad_norm": 1.2811965942382812, |
| "learning_rate": 7.044548299961412e-06, |
| "loss": 3.1629, |
| "step": 552000 |
| }, |
| { |
| "epoch": 23.289634531888883, |
| "grad_norm": 1.2909471988677979, |
| "learning_rate": 6.958796038245509e-06, |
| "loss": 3.1541, |
| "step": 552500 |
| }, |
| { |
| "epoch": 23.310711124225435, |
| "grad_norm": 1.3470662832260132, |
| "learning_rate": 6.873043776529606e-06, |
| "loss": 3.1632, |
| "step": 553000 |
| }, |
| { |
| "epoch": 23.331787716561987, |
| "grad_norm": 1.2972832918167114, |
| "learning_rate": 6.787291514813704e-06, |
| "loss": 3.156, |
| "step": 553500 |
| }, |
| { |
| "epoch": 23.35286430889854, |
| "grad_norm": 1.291892409324646, |
| "learning_rate": 6.701539253097801e-06, |
| "loss": 3.1583, |
| "step": 554000 |
| }, |
| { |
| "epoch": 23.373940901235088, |
| "grad_norm": 1.2642675638198853, |
| "learning_rate": 6.615786991381898e-06, |
| "loss": 3.1583, |
| "step": 554500 |
| }, |
| { |
| "epoch": 23.39501749357164, |
| "grad_norm": 1.2373483180999756, |
| "learning_rate": 6.530034729665996e-06, |
| "loss": 3.1583, |
| "step": 555000 |
| }, |
| { |
| "epoch": 23.416094085908192, |
| "grad_norm": 1.2409842014312744, |
| "learning_rate": 6.444282467950093e-06, |
| "loss": 3.1571, |
| "step": 555500 |
| }, |
| { |
| "epoch": 23.43717067824474, |
| "grad_norm": 1.2378615140914917, |
| "learning_rate": 6.358530206234189e-06, |
| "loss": 3.1546, |
| "step": 556000 |
| }, |
| { |
| "epoch": 23.458247270581293, |
| "grad_norm": 1.254681944847107, |
| "learning_rate": 6.272777944518287e-06, |
| "loss": 3.1604, |
| "step": 556500 |
| }, |
| { |
| "epoch": 23.479323862917845, |
| "grad_norm": 1.2594963312149048, |
| "learning_rate": 6.187025682802384e-06, |
| "loss": 3.1596, |
| "step": 557000 |
| }, |
| { |
| "epoch": 23.500400455254393, |
| "grad_norm": 1.2599483728408813, |
| "learning_rate": 6.101273421086482e-06, |
| "loss": 3.1575, |
| "step": 557500 |
| }, |
| { |
| "epoch": 23.521477047590945, |
| "grad_norm": 1.2755753993988037, |
| "learning_rate": 6.015521159370579e-06, |
| "loss": 3.1576, |
| "step": 558000 |
| }, |
| { |
| "epoch": 23.542553639927497, |
| "grad_norm": 1.2996673583984375, |
| "learning_rate": 5.929768897654676e-06, |
| "loss": 3.1565, |
| "step": 558500 |
| }, |
| { |
| "epoch": 23.563630232264046, |
| "grad_norm": 1.2656636238098145, |
| "learning_rate": 5.844016635938773e-06, |
| "loss": 3.1578, |
| "step": 559000 |
| }, |
| { |
| "epoch": 23.584706824600598, |
| "grad_norm": 1.2176569700241089, |
| "learning_rate": 5.75826437422287e-06, |
| "loss": 3.1566, |
| "step": 559500 |
| }, |
| { |
| "epoch": 23.60578341693715, |
| "grad_norm": 1.2881158590316772, |
| "learning_rate": 5.672512112506967e-06, |
| "loss": 3.1579, |
| "step": 560000 |
| }, |
| { |
| "epoch": 23.60578341693715, |
| "eval_accuracy": 0.7603135556989984, |
| "eval_loss": 2.9512956142425537, |
| "eval_runtime": 181.8817, |
| "eval_samples_per_second": 663.651, |
| "eval_steps_per_second": 6.917, |
| "step": 560000 |
| }, |
| { |
| "epoch": 23.6268600092737, |
| "grad_norm": 1.2816041707992554, |
| "learning_rate": 5.5867598507910655e-06, |
| "loss": 3.1578, |
| "step": 560500 |
| }, |
| { |
| "epoch": 23.64793660161025, |
| "grad_norm": 1.2905371189117432, |
| "learning_rate": 5.501007589075162e-06, |
| "loss": 3.1574, |
| "step": 561000 |
| }, |
| { |
| "epoch": 23.669013193946803, |
| "grad_norm": 1.3094907999038696, |
| "learning_rate": 5.41525532735926e-06, |
| "loss": 3.1593, |
| "step": 561500 |
| }, |
| { |
| "epoch": 23.690089786283355, |
| "grad_norm": 1.3383922576904297, |
| "learning_rate": 5.329503065643356e-06, |
| "loss": 3.1535, |
| "step": 562000 |
| }, |
| { |
| "epoch": 23.711166378619904, |
| "grad_norm": 1.286407709121704, |
| "learning_rate": 5.243750803927454e-06, |
| "loss": 3.1541, |
| "step": 562500 |
| }, |
| { |
| "epoch": 23.732242970956456, |
| "grad_norm": 1.3089640140533447, |
| "learning_rate": 5.157998542211551e-06, |
| "loss": 3.1572, |
| "step": 563000 |
| }, |
| { |
| "epoch": 23.753319563293008, |
| "grad_norm": 1.2453045845031738, |
| "learning_rate": 5.072246280495649e-06, |
| "loss": 3.1541, |
| "step": 563500 |
| }, |
| { |
| "epoch": 23.774396155629557, |
| "grad_norm": 1.2313088178634644, |
| "learning_rate": 4.986494018779746e-06, |
| "loss": 3.1569, |
| "step": 564000 |
| }, |
| { |
| "epoch": 23.79547274796611, |
| "grad_norm": 1.282601237297058, |
| "learning_rate": 4.900741757063843e-06, |
| "loss": 3.1551, |
| "step": 564500 |
| }, |
| { |
| "epoch": 23.81654934030266, |
| "grad_norm": 1.459593653678894, |
| "learning_rate": 4.81498949534794e-06, |
| "loss": 3.1545, |
| "step": 565000 |
| }, |
| { |
| "epoch": 23.83762593263921, |
| "grad_norm": 1.2445076704025269, |
| "learning_rate": 4.729237233632038e-06, |
| "loss": 3.1548, |
| "step": 565500 |
| }, |
| { |
| "epoch": 23.85870252497576, |
| "grad_norm": 1.2471879720687866, |
| "learning_rate": 4.643484971916134e-06, |
| "loss": 3.1554, |
| "step": 566000 |
| }, |
| { |
| "epoch": 23.879779117312314, |
| "grad_norm": 1.2625129222869873, |
| "learning_rate": 4.557732710200232e-06, |
| "loss": 3.1541, |
| "step": 566500 |
| }, |
| { |
| "epoch": 23.900855709648862, |
| "grad_norm": 1.27744722366333, |
| "learning_rate": 4.471980448484329e-06, |
| "loss": 3.1557, |
| "step": 567000 |
| }, |
| { |
| "epoch": 23.921932301985414, |
| "grad_norm": 1.278097152709961, |
| "learning_rate": 4.386228186768427e-06, |
| "loss": 3.1539, |
| "step": 567500 |
| }, |
| { |
| "epoch": 23.943008894321967, |
| "grad_norm": 1.2704708576202393, |
| "learning_rate": 4.300475925052524e-06, |
| "loss": 3.1551, |
| "step": 568000 |
| }, |
| { |
| "epoch": 23.96408548665852, |
| "grad_norm": 1.2700526714324951, |
| "learning_rate": 4.214723663336621e-06, |
| "loss": 3.158, |
| "step": 568500 |
| }, |
| { |
| "epoch": 23.985162078995067, |
| "grad_norm": 1.295990228652954, |
| "learning_rate": 4.128971401620718e-06, |
| "loss": 3.1536, |
| "step": 569000 |
| }, |
| { |
| "epoch": 24.00623867133162, |
| "grad_norm": 1.2671464681625366, |
| "learning_rate": 4.043219139904815e-06, |
| "loss": 3.1548, |
| "step": 569500 |
| }, |
| { |
| "epoch": 24.02731526366817, |
| "grad_norm": 1.2454357147216797, |
| "learning_rate": 3.957466878188912e-06, |
| "loss": 3.154, |
| "step": 570000 |
| }, |
| { |
| "epoch": 24.02731526366817, |
| "eval_accuracy": 0.76052697368067, |
| "eval_loss": 2.9501137733459473, |
| "eval_runtime": 179.6607, |
| "eval_samples_per_second": 671.855, |
| "eval_steps_per_second": 7.002, |
| "step": 570000 |
| }, |
| { |
| "epoch": 24.04839185600472, |
| "grad_norm": 1.2714110612869263, |
| "learning_rate": 3.87171461647301e-06, |
| "loss": 3.1575, |
| "step": 570500 |
| }, |
| { |
| "epoch": 24.069468448341272, |
| "grad_norm": 1.2464501857757568, |
| "learning_rate": 3.7859623547571067e-06, |
| "loss": 3.1511, |
| "step": 571000 |
| }, |
| { |
| "epoch": 24.090545040677824, |
| "grad_norm": 1.244425654411316, |
| "learning_rate": 3.7002100930412046e-06, |
| "loss": 3.1492, |
| "step": 571500 |
| }, |
| { |
| "epoch": 24.111621633014373, |
| "grad_norm": 1.2312793731689453, |
| "learning_rate": 3.614457831325301e-06, |
| "loss": 3.1567, |
| "step": 572000 |
| }, |
| { |
| "epoch": 24.132698225350925, |
| "grad_norm": 1.3055317401885986, |
| "learning_rate": 3.5287055696093987e-06, |
| "loss": 3.153, |
| "step": 572500 |
| }, |
| { |
| "epoch": 24.153774817687477, |
| "grad_norm": 1.2372642755508423, |
| "learning_rate": 3.4429533078934957e-06, |
| "loss": 3.1548, |
| "step": 573000 |
| }, |
| { |
| "epoch": 24.174851410024026, |
| "grad_norm": 1.2717217206954956, |
| "learning_rate": 3.357201046177593e-06, |
| "loss": 3.1551, |
| "step": 573500 |
| }, |
| { |
| "epoch": 24.195928002360578, |
| "grad_norm": 1.2693278789520264, |
| "learning_rate": 3.2714487844616906e-06, |
| "loss": 3.1541, |
| "step": 574000 |
| }, |
| { |
| "epoch": 24.21700459469713, |
| "grad_norm": 1.3049947023391724, |
| "learning_rate": 3.1856965227457872e-06, |
| "loss": 3.154, |
| "step": 574500 |
| }, |
| { |
| "epoch": 24.238081187033682, |
| "grad_norm": 1.267085075378418, |
| "learning_rate": 3.099944261029885e-06, |
| "loss": 3.1583, |
| "step": 575000 |
| }, |
| { |
| "epoch": 24.25915777937023, |
| "grad_norm": 1.2947229146957397, |
| "learning_rate": 3.014191999313982e-06, |
| "loss": 3.1525, |
| "step": 575500 |
| }, |
| { |
| "epoch": 24.280234371706783, |
| "grad_norm": 1.2750862836837769, |
| "learning_rate": 2.928439737598079e-06, |
| "loss": 3.1509, |
| "step": 576000 |
| }, |
| { |
| "epoch": 24.301310964043335, |
| "grad_norm": 1.2644596099853516, |
| "learning_rate": 2.8426874758821767e-06, |
| "loss": 3.1534, |
| "step": 576500 |
| }, |
| { |
| "epoch": 24.322387556379883, |
| "grad_norm": 1.2458935976028442, |
| "learning_rate": 2.7569352141662737e-06, |
| "loss": 3.1493, |
| "step": 577000 |
| }, |
| { |
| "epoch": 24.343464148716436, |
| "grad_norm": 1.2863774299621582, |
| "learning_rate": 2.6711829524503707e-06, |
| "loss": 3.1497, |
| "step": 577500 |
| }, |
| { |
| "epoch": 24.364540741052988, |
| "grad_norm": 1.2988264560699463, |
| "learning_rate": 2.585430690734468e-06, |
| "loss": 3.1498, |
| "step": 578000 |
| }, |
| { |
| "epoch": 24.385617333389536, |
| "grad_norm": 1.3064343929290771, |
| "learning_rate": 2.4996784290185657e-06, |
| "loss": 3.1516, |
| "step": 578500 |
| }, |
| { |
| "epoch": 24.40669392572609, |
| "grad_norm": 1.2625620365142822, |
| "learning_rate": 2.4139261673026627e-06, |
| "loss": 3.1487, |
| "step": 579000 |
| }, |
| { |
| "epoch": 24.42777051806264, |
| "grad_norm": 1.265486478805542, |
| "learning_rate": 2.3281739055867597e-06, |
| "loss": 3.1517, |
| "step": 579500 |
| }, |
| { |
| "epoch": 24.44884711039919, |
| "grad_norm": 1.2586462497711182, |
| "learning_rate": 2.242421643870857e-06, |
| "loss": 3.1504, |
| "step": 580000 |
| }, |
| { |
| "epoch": 24.44884711039919, |
| "eval_accuracy": 0.7609636478606847, |
| "eval_loss": 2.946443557739258, |
| "eval_runtime": 182.3247, |
| "eval_samples_per_second": 662.039, |
| "eval_steps_per_second": 6.9, |
| "step": 580000 |
| }, |
| { |
| "epoch": 24.46992370273574, |
| "grad_norm": 1.287256121635437, |
| "learning_rate": 2.1566693821549547e-06, |
| "loss": 3.1509, |
| "step": 580500 |
| }, |
| { |
| "epoch": 24.491000295072293, |
| "grad_norm": 1.2539594173431396, |
| "learning_rate": 2.0709171204390517e-06, |
| "loss": 3.155, |
| "step": 581000 |
| }, |
| { |
| "epoch": 24.512076887408845, |
| "grad_norm": 1.304165244102478, |
| "learning_rate": 1.985164858723149e-06, |
| "loss": 3.1492, |
| "step": 581500 |
| }, |
| { |
| "epoch": 24.533153479745394, |
| "grad_norm": 1.2802070379257202, |
| "learning_rate": 1.8994125970072462e-06, |
| "loss": 3.1543, |
| "step": 582000 |
| }, |
| { |
| "epoch": 24.554230072081946, |
| "grad_norm": 1.2618753910064697, |
| "learning_rate": 1.8136603352913432e-06, |
| "loss": 3.148, |
| "step": 582500 |
| }, |
| { |
| "epoch": 24.5753066644185, |
| "grad_norm": 1.2322430610656738, |
| "learning_rate": 1.7279080735754405e-06, |
| "loss": 3.1534, |
| "step": 583000 |
| }, |
| { |
| "epoch": 24.596383256755047, |
| "grad_norm": 1.2677052021026611, |
| "learning_rate": 1.642155811859538e-06, |
| "loss": 3.1534, |
| "step": 583500 |
| }, |
| { |
| "epoch": 24.6174598490916, |
| "grad_norm": 1.2846912145614624, |
| "learning_rate": 1.5564035501436352e-06, |
| "loss": 3.1509, |
| "step": 584000 |
| }, |
| { |
| "epoch": 24.63853644142815, |
| "grad_norm": 1.266931414604187, |
| "learning_rate": 1.4706512884277324e-06, |
| "loss": 3.1516, |
| "step": 584500 |
| }, |
| { |
| "epoch": 24.6596130337647, |
| "grad_norm": 1.2346444129943848, |
| "learning_rate": 1.3848990267118295e-06, |
| "loss": 3.1486, |
| "step": 585000 |
| }, |
| { |
| "epoch": 24.680689626101252, |
| "grad_norm": 1.2472306489944458, |
| "learning_rate": 1.299146764995927e-06, |
| "loss": 3.1499, |
| "step": 585500 |
| }, |
| { |
| "epoch": 24.701766218437804, |
| "grad_norm": 1.3008043766021729, |
| "learning_rate": 1.213394503280024e-06, |
| "loss": 3.1495, |
| "step": 586000 |
| }, |
| { |
| "epoch": 24.722842810774353, |
| "grad_norm": 1.2973970174789429, |
| "learning_rate": 1.1276422415641212e-06, |
| "loss": 3.1495, |
| "step": 586500 |
| }, |
| { |
| "epoch": 24.743919403110905, |
| "grad_norm": 1.2528910636901855, |
| "learning_rate": 1.0418899798482187e-06, |
| "loss": 3.1513, |
| "step": 587000 |
| }, |
| { |
| "epoch": 24.764995995447457, |
| "grad_norm": 1.2960785627365112, |
| "learning_rate": 9.561377181323157e-07, |
| "loss": 3.1482, |
| "step": 587500 |
| }, |
| { |
| "epoch": 24.786072587784005, |
| "grad_norm": 1.2799756526947021, |
| "learning_rate": 8.70385456416413e-07, |
| "loss": 3.1496, |
| "step": 588000 |
| }, |
| { |
| "epoch": 24.807149180120557, |
| "grad_norm": 1.242201328277588, |
| "learning_rate": 7.846331947005103e-07, |
| "loss": 3.1509, |
| "step": 588500 |
| }, |
| { |
| "epoch": 24.82822577245711, |
| "grad_norm": 1.2769993543624878, |
| "learning_rate": 6.988809329846075e-07, |
| "loss": 3.1468, |
| "step": 589000 |
| }, |
| { |
| "epoch": 24.84930236479366, |
| "grad_norm": 1.2839336395263672, |
| "learning_rate": 6.131286712687047e-07, |
| "loss": 3.1518, |
| "step": 589500 |
| }, |
| { |
| "epoch": 24.87037895713021, |
| "grad_norm": 1.2706698179244995, |
| "learning_rate": 5.27376409552802e-07, |
| "loss": 3.1482, |
| "step": 590000 |
| }, |
| { |
| "epoch": 24.87037895713021, |
| "eval_accuracy": 0.7609835632187141, |
| "eval_loss": 2.9464175701141357, |
| "eval_runtime": 182.9226, |
| "eval_samples_per_second": 659.875, |
| "eval_steps_per_second": 6.877, |
| "step": 590000 |
| }, |
| { |
| "epoch": 24.891455549466762, |
| "grad_norm": 1.2449679374694824, |
| "learning_rate": 4.4162414783689917e-07, |
| "loss": 3.1509, |
| "step": 590500 |
| }, |
| { |
| "epoch": 24.912532141803315, |
| "grad_norm": 1.251673936843872, |
| "learning_rate": 3.558718861209964e-07, |
| "loss": 3.1481, |
| "step": 591000 |
| }, |
| { |
| "epoch": 24.933608734139863, |
| "grad_norm": 1.268472671508789, |
| "learning_rate": 2.701196244050937e-07, |
| "loss": 3.1494, |
| "step": 591500 |
| }, |
| { |
| "epoch": 24.954685326476415, |
| "grad_norm": 1.3184300661087036, |
| "learning_rate": 1.8436736268919094e-07, |
| "loss": 3.1499, |
| "step": 592000 |
| }, |
| { |
| "epoch": 24.975761918812967, |
| "grad_norm": 1.2411932945251465, |
| "learning_rate": 9.861510097328816e-08, |
| "loss": 3.1489, |
| "step": 592500 |
| }, |
| { |
| "epoch": 24.996838511149516, |
| "grad_norm": 1.2670942544937134, |
| "learning_rate": 1.2862839257385414e-08, |
| "loss": 3.1495, |
| "step": 593000 |
| }, |
| { |
| "epoch": 25.0, |
| "step": 593075, |
| "total_flos": 1.508982381118034e+19, |
| "train_loss": 3.6999233157054943, |
| "train_runtime": 196537.7739, |
| "train_samples_per_second": 289.682, |
| "train_steps_per_second": 3.018 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 593075, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 25, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.508982381118034e+19, |
| "train_batch_size": 96, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
