| { |
| "best_metric": 3.3065290451049805, |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_634/checkpoint-90000", |
| "epoch": 10.0, |
| "eval_steps": 1000, |
| "global_step": 92910, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.005381552039608223, |
| "grad_norm": 2.728989362716675, |
| "learning_rate": 0.00028199999999999997, |
| "loss": 8.6119, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.010763104079216447, |
| "grad_norm": 1.724418044090271, |
| "learning_rate": 0.0005819999999999999, |
| "loss": 6.9596, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.01614465611882467, |
| "grad_norm": 1.106858491897583, |
| "learning_rate": 0.0005996961534317422, |
| "loss": 6.5223, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.021526208158432893, |
| "grad_norm": 1.3945527076721191, |
| "learning_rate": 0.0005993729124016807, |
| "loss": 6.2759, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.026907760198041114, |
| "grad_norm": 1.685577630996704, |
| "learning_rate": 0.0005990496713716194, |
| "loss": 6.0955, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.03228931223764934, |
| "grad_norm": 2.0357065200805664, |
| "learning_rate": 0.000598726430341558, |
| "loss": 6.0263, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03767086427725756, |
| "grad_norm": 0.9172292351722717, |
| "learning_rate": 0.0005984031893114966, |
| "loss": 5.8966, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.04305241631686579, |
| "grad_norm": 2.3104074001312256, |
| "learning_rate": 0.0005980799482814351, |
| "loss": 5.8365, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.048433968356474004, |
| "grad_norm": 1.7049633264541626, |
| "learning_rate": 0.0005977567072513737, |
| "loss": 5.7381, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.05381552039608223, |
| "grad_norm": 1.1454957723617554, |
| "learning_rate": 0.0005974334662213123, |
| "loss": 5.6778, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.05919707243569045, |
| "grad_norm": 1.5053794384002686, |
| "learning_rate": 0.0005971102251912509, |
| "loss": 5.6187, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.06457862447529868, |
| "grad_norm": 0.9626969695091248, |
| "learning_rate": 0.0005967869841611895, |
| "loss": 5.5314, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.0699601765149069, |
| "grad_norm": 1.2051186561584473, |
| "learning_rate": 0.000596463743131128, |
| "loss": 5.458, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.07534172855451512, |
| "grad_norm": 1.3867778778076172, |
| "learning_rate": 0.0005961405021010667, |
| "loss": 5.3855, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.08072328059412334, |
| "grad_norm": 1.4545854330062866, |
| "learning_rate": 0.0005958172610710052, |
| "loss": 5.3445, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.08610483263373157, |
| "grad_norm": 1.1404812335968018, |
| "learning_rate": 0.0005954940200409439, |
| "loss": 5.2568, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.09148638467333979, |
| "grad_norm": 1.3057512044906616, |
| "learning_rate": 0.0005951707790108824, |
| "loss": 5.2444, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.09686793671294801, |
| "grad_norm": 1.1935104131698608, |
| "learning_rate": 0.0005948475379808209, |
| "loss": 5.1903, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.10224948875255624, |
| "grad_norm": 0.9733734130859375, |
| "learning_rate": 0.0005945242969507596, |
| "loss": 5.1723, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.10763104079216446, |
| "grad_norm": 1.0952446460723877, |
| "learning_rate": 0.0005942010559206981, |
| "loss": 5.1166, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.10763104079216446, |
| "eval_accuracy": 0.22600756810824782, |
| "eval_loss": 5.031754493713379, |
| "eval_runtime": 180.4654, |
| "eval_samples_per_second": 99.803, |
| "eval_steps_per_second": 6.239, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.11301259283177269, |
| "grad_norm": 1.521315097808838, |
| "learning_rate": 0.0005938778148906367, |
| "loss": 5.0458, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.1183941448713809, |
| "grad_norm": 1.205197811126709, |
| "learning_rate": 0.0005935545738605753, |
| "loss": 5.023, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.12377569691098914, |
| "grad_norm": 0.8693559169769287, |
| "learning_rate": 0.000593231332830514, |
| "loss": 4.9911, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.12915724895059735, |
| "grad_norm": 1.2472184896469116, |
| "learning_rate": 0.0005929080918004525, |
| "loss": 4.9623, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.13453880099020557, |
| "grad_norm": 0.9657862186431885, |
| "learning_rate": 0.000592584850770391, |
| "loss": 4.9466, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.1399203530298138, |
| "grad_norm": 0.8568148016929626, |
| "learning_rate": 0.0005922616097403296, |
| "loss": 4.9032, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.14530190506942203, |
| "grad_norm": 0.949377179145813, |
| "learning_rate": 0.0005919383687102682, |
| "loss": 4.8553, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.15068345710903025, |
| "grad_norm": 0.92209392786026, |
| "learning_rate": 0.0005916151276802069, |
| "loss": 4.8447, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.15606500914863847, |
| "grad_norm": 0.7755404710769653, |
| "learning_rate": 0.0005912918866501454, |
| "loss": 4.8361, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.16144656118824668, |
| "grad_norm": 1.0212815999984741, |
| "learning_rate": 0.000590968645620084, |
| "loss": 4.7901, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.1668281132278549, |
| "grad_norm": 0.9886590838432312, |
| "learning_rate": 0.0005906454045900226, |
| "loss": 4.7598, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.17220966526746315, |
| "grad_norm": 1.1899733543395996, |
| "learning_rate": 0.0005903221635599612, |
| "loss": 4.7607, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.17759121730707136, |
| "grad_norm": 0.948335587978363, |
| "learning_rate": 0.0005899989225298998, |
| "loss": 4.7664, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.18297276934667958, |
| "grad_norm": 1.285933256149292, |
| "learning_rate": 0.0005896756814998383, |
| "loss": 4.7184, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.1883543213862878, |
| "grad_norm": 0.788101315498352, |
| "learning_rate": 0.0005893524404697769, |
| "loss": 4.6973, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.19373587342589602, |
| "grad_norm": 0.6696946620941162, |
| "learning_rate": 0.0005890291994397155, |
| "loss": 4.6636, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.19911742546550426, |
| "grad_norm": 0.7915722727775574, |
| "learning_rate": 0.0005887059584096541, |
| "loss": 4.6548, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.20449897750511248, |
| "grad_norm": 0.7558673620223999, |
| "learning_rate": 0.0005883827173795926, |
| "loss": 4.6091, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.2098805295447207, |
| "grad_norm": 0.9301922917366028, |
| "learning_rate": 0.0005880594763495313, |
| "loss": 4.6026, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.2152620815843289, |
| "grad_norm": 1.3686238527297974, |
| "learning_rate": 0.0005877362353194698, |
| "loss": 4.5936, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.2152620815843289, |
| "eval_accuracy": 0.2692196908151694, |
| "eval_loss": 4.520735263824463, |
| "eval_runtime": 180.5703, |
| "eval_samples_per_second": 99.745, |
| "eval_steps_per_second": 6.236, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.22064363362393713, |
| "grad_norm": 0.808482825756073, |
| "learning_rate": 0.0005874129942894084, |
| "loss": 4.5665, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.22602518566354537, |
| "grad_norm": 0.9632678031921387, |
| "learning_rate": 0.000587089753259347, |
| "loss": 4.5801, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.2314067377031536, |
| "grad_norm": 0.8442896604537964, |
| "learning_rate": 0.0005867665122292855, |
| "loss": 4.5534, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.2367882897427618, |
| "grad_norm": 1.0314425230026245, |
| "learning_rate": 0.0005864432711992242, |
| "loss": 4.5134, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.24216984178237003, |
| "grad_norm": 1.247391700744629, |
| "learning_rate": 0.0005861200301691628, |
| "loss": 4.4984, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.24755139382197827, |
| "grad_norm": 0.8239965438842773, |
| "learning_rate": 0.0005857967891391014, |
| "loss": 4.4888, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.2529329458615865, |
| "grad_norm": 0.69472736120224, |
| "learning_rate": 0.0005854735481090399, |
| "loss": 4.4592, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.2583144979011947, |
| "grad_norm": 1.0152015686035156, |
| "learning_rate": 0.0005851503070789784, |
| "loss": 4.4467, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2636960499408029, |
| "grad_norm": 0.9481901526451111, |
| "learning_rate": 0.0005848270660489171, |
| "loss": 4.4431, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.26907760198041114, |
| "grad_norm": 0.914855420589447, |
| "learning_rate": 0.0005845038250188556, |
| "loss": 4.4427, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.27445915402001936, |
| "grad_norm": 0.7849532961845398, |
| "learning_rate": 0.0005841805839887943, |
| "loss": 4.4114, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.2798407060596276, |
| "grad_norm": 0.8744374513626099, |
| "learning_rate": 0.0005838573429587328, |
| "loss": 4.3739, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2852222580992358, |
| "grad_norm": 0.8036152124404907, |
| "learning_rate": 0.0005835341019286715, |
| "loss": 4.4008, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.29060381013884407, |
| "grad_norm": 0.7502331137657166, |
| "learning_rate": 0.00058321086089861, |
| "loss": 4.361, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.2959853621784523, |
| "grad_norm": 0.737133800983429, |
| "learning_rate": 0.0005828876198685486, |
| "loss": 4.3553, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.3013669142180605, |
| "grad_norm": 0.90378338098526, |
| "learning_rate": 0.0005825643788384872, |
| "loss": 4.3844, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.3067484662576687, |
| "grad_norm": 0.6965075135231018, |
| "learning_rate": 0.0005822411378084257, |
| "loss": 4.3413, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.31213001829727693, |
| "grad_norm": 0.8999699354171753, |
| "learning_rate": 0.0005819178967783644, |
| "loss": 4.3424, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.31751157033688515, |
| "grad_norm": 0.7911924719810486, |
| "learning_rate": 0.0005815946557483029, |
| "loss": 4.3326, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.32289312237649337, |
| "grad_norm": 1.0059629678726196, |
| "learning_rate": 0.0005812714147182415, |
| "loss": 4.3077, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.32289312237649337, |
| "eval_accuracy": 0.29792002517270105, |
| "eval_loss": 4.243895530700684, |
| "eval_runtime": 180.4537, |
| "eval_samples_per_second": 99.81, |
| "eval_steps_per_second": 6.24, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.3282746744161016, |
| "grad_norm": 0.7331672310829163, |
| "learning_rate": 0.0005809481736881801, |
| "loss": 4.3384, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.3336562264557098, |
| "grad_norm": 0.9308981895446777, |
| "learning_rate": 0.0005806249326581187, |
| "loss": 4.2823, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.3390377784953181, |
| "grad_norm": 0.7922755479812622, |
| "learning_rate": 0.0005803016916280573, |
| "loss": 4.3038, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.3444193305349263, |
| "grad_norm": 0.8320370316505432, |
| "learning_rate": 0.0005799784505979959, |
| "loss": 4.2708, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.3498008825745345, |
| "grad_norm": 0.7593258023262024, |
| "learning_rate": 0.0005796552095679344, |
| "loss": 4.2699, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.35518243461414273, |
| "grad_norm": 0.8462849259376526, |
| "learning_rate": 0.000579331968537873, |
| "loss": 4.2502, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.36056398665375095, |
| "grad_norm": 0.6994985938072205, |
| "learning_rate": 0.0005790087275078116, |
| "loss": 4.2565, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.36594553869335916, |
| "grad_norm": 0.7432887554168701, |
| "learning_rate": 0.0005786854864777502, |
| "loss": 4.2457, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.3713270907329674, |
| "grad_norm": 0.7776950597763062, |
| "learning_rate": 0.0005783622454476888, |
| "loss": 4.2194, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.3767086427725756, |
| "grad_norm": 0.769761323928833, |
| "learning_rate": 0.0005780390044176273, |
| "loss": 4.2308, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.3820901948121838, |
| "grad_norm": 0.739103376865387, |
| "learning_rate": 0.000577715763387566, |
| "loss": 4.2353, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.38747174685179203, |
| "grad_norm": 0.881566047668457, |
| "learning_rate": 0.0005773925223575045, |
| "loss": 4.1979, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.3928532988914003, |
| "grad_norm": 0.7779421210289001, |
| "learning_rate": 0.0005770692813274432, |
| "loss": 4.2024, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.3982348509310085, |
| "grad_norm": 0.8221611380577087, |
| "learning_rate": 0.0005767460402973817, |
| "loss": 4.2008, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.40361640297061674, |
| "grad_norm": 0.749286949634552, |
| "learning_rate": 0.0005764227992673203, |
| "loss": 4.1936, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.40899795501022496, |
| "grad_norm": 0.6680869460105896, |
| "learning_rate": 0.0005760995582372589, |
| "loss": 4.1999, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.4143795070498332, |
| "grad_norm": 0.650600254535675, |
| "learning_rate": 0.0005757763172071974, |
| "loss": 4.1712, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.4197610590894414, |
| "grad_norm": 0.7819272875785828, |
| "learning_rate": 0.000575453076177136, |
| "loss": 4.1907, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.4251426111290496, |
| "grad_norm": 0.8200270533561707, |
| "learning_rate": 0.0005751298351470746, |
| "loss": 4.1628, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.4305241631686578, |
| "grad_norm": 0.5963762998580933, |
| "learning_rate": 0.0005748065941170133, |
| "loss": 4.1514, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.4305241631686578, |
| "eval_accuracy": 0.31238215915235973, |
| "eval_loss": 4.085026741027832, |
| "eval_runtime": 180.1755, |
| "eval_samples_per_second": 99.964, |
| "eval_steps_per_second": 6.249, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.43590571520826604, |
| "grad_norm": 0.7472826838493347, |
| "learning_rate": 0.0005744833530869518, |
| "loss": 4.1579, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.44128726724787426, |
| "grad_norm": 0.5647183656692505, |
| "learning_rate": 0.0005741601120568903, |
| "loss": 4.153, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.44666881928748253, |
| "grad_norm": 0.6447238922119141, |
| "learning_rate": 0.0005738368710268289, |
| "loss": 4.1402, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.45205037132709075, |
| "grad_norm": 0.6352419257164001, |
| "learning_rate": 0.0005735136299967675, |
| "loss": 4.1218, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.45743192336669897, |
| "grad_norm": 0.7383701205253601, |
| "learning_rate": 0.0005731903889667062, |
| "loss": 4.1315, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.4628134754063072, |
| "grad_norm": 0.6032237410545349, |
| "learning_rate": 0.0005728671479366447, |
| "loss": 4.1318, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.4681950274459154, |
| "grad_norm": 0.7067080736160278, |
| "learning_rate": 0.0005725439069065833, |
| "loss": 4.1231, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.4735765794855236, |
| "grad_norm": 0.6998112797737122, |
| "learning_rate": 0.0005722206658765219, |
| "loss": 4.1332, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.47895813152513184, |
| "grad_norm": 0.6890502572059631, |
| "learning_rate": 0.0005718974248464605, |
| "loss": 4.0919, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.48433968356474005, |
| "grad_norm": 0.6913346648216248, |
| "learning_rate": 0.000571574183816399, |
| "loss": 4.0985, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.48972123560434827, |
| "grad_norm": 0.6454923748970032, |
| "learning_rate": 0.0005712509427863376, |
| "loss": 4.0914, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.49510278764395654, |
| "grad_norm": 0.6718611717224121, |
| "learning_rate": 0.0005709277017562762, |
| "loss": 4.0875, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.5004843396835648, |
| "grad_norm": 0.6115706562995911, |
| "learning_rate": 0.0005706044607262148, |
| "loss": 4.0789, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.505865891723173, |
| "grad_norm": 0.7557240724563599, |
| "learning_rate": 0.0005702812196961534, |
| "loss": 4.1053, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.5112474437627812, |
| "grad_norm": 0.6292441487312317, |
| "learning_rate": 0.0005699579786660919, |
| "loss": 4.0814, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.5166289958023894, |
| "grad_norm": 0.5616717338562012, |
| "learning_rate": 0.0005696347376360306, |
| "loss": 4.0577, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.5220105478419976, |
| "grad_norm": 0.5930373668670654, |
| "learning_rate": 0.0005693114966059691, |
| "loss": 4.0565, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.5273920998816058, |
| "grad_norm": 0.7059026956558228, |
| "learning_rate": 0.0005689882555759077, |
| "loss": 4.0902, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.5327736519212141, |
| "grad_norm": 0.6343038082122803, |
| "learning_rate": 0.0005686650145458463, |
| "loss": 4.0649, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.5381552039608223, |
| "grad_norm": 0.5513986945152283, |
| "learning_rate": 0.0005683417735157848, |
| "loss": 4.049, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5381552039608223, |
| "eval_accuracy": 0.3214264255340371, |
| "eval_loss": 3.9884958267211914, |
| "eval_runtime": 180.1991, |
| "eval_samples_per_second": 99.951, |
| "eval_steps_per_second": 6.249, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5435367560004305, |
| "grad_norm": 0.6979557871818542, |
| "learning_rate": 0.0005680185324857235, |
| "loss": 4.0494, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.5489183080400387, |
| "grad_norm": 0.7719425559043884, |
| "learning_rate": 0.0005676952914556621, |
| "loss": 4.0467, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.5542998600796469, |
| "grad_norm": 0.623746931552887, |
| "learning_rate": 0.0005673720504256007, |
| "loss": 4.0159, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.5596814121192552, |
| "grad_norm": 0.694391667842865, |
| "learning_rate": 0.0005670488093955392, |
| "loss": 4.0426, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.5650629641588634, |
| "grad_norm": 0.6758689284324646, |
| "learning_rate": 0.0005667255683654777, |
| "loss": 4.0391, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.5704445161984716, |
| "grad_norm": 0.6376744508743286, |
| "learning_rate": 0.0005664023273354164, |
| "loss": 4.0361, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.5758260682380799, |
| "grad_norm": 0.6066127419471741, |
| "learning_rate": 0.0005660790863053549, |
| "loss": 4.0327, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.5812076202776881, |
| "grad_norm": 0.5952569246292114, |
| "learning_rate": 0.0005657558452752936, |
| "loss": 4.0252, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.5865891723172963, |
| "grad_norm": 0.6313889026641846, |
| "learning_rate": 0.0005654326042452321, |
| "loss": 4.0221, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.5919707243569046, |
| "grad_norm": 0.596638023853302, |
| "learning_rate": 0.0005651093632151708, |
| "loss": 4.0425, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.5973522763965128, |
| "grad_norm": 0.6534123420715332, |
| "learning_rate": 0.0005647861221851093, |
| "loss": 4.0383, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.602733828436121, |
| "grad_norm": 0.5690531730651855, |
| "learning_rate": 0.0005644628811550479, |
| "loss": 4.0233, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.6081153804757292, |
| "grad_norm": 0.626017153263092, |
| "learning_rate": 0.0005641396401249865, |
| "loss": 3.9948, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.6134969325153374, |
| "grad_norm": 0.6084668636322021, |
| "learning_rate": 0.000563816399094925, |
| "loss": 4.0122, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.6188784845549457, |
| "grad_norm": 0.5626436471939087, |
| "learning_rate": 0.0005634931580648637, |
| "loss": 4.0019, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.6242600365945539, |
| "grad_norm": 0.5933359861373901, |
| "learning_rate": 0.0005631699170348022, |
| "loss": 3.9866, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.6296415886341621, |
| "grad_norm": 0.5561618804931641, |
| "learning_rate": 0.0005628466760047408, |
| "loss": 3.9969, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.6350231406737703, |
| "grad_norm": 0.6588941812515259, |
| "learning_rate": 0.0005625234349746794, |
| "loss": 3.9841, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.6404046927133785, |
| "grad_norm": 0.6015214920043945, |
| "learning_rate": 0.000562200193944618, |
| "loss": 3.9915, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.6457862447529867, |
| "grad_norm": 0.6395725607872009, |
| "learning_rate": 0.0005618769529145566, |
| "loss": 3.9874, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6457862447529867, |
| "eval_accuracy": 0.32797471777684023, |
| "eval_loss": 3.916991710662842, |
| "eval_runtime": 180.0235, |
| "eval_samples_per_second": 100.048, |
| "eval_steps_per_second": 6.255, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.651167796792595, |
| "grad_norm": 0.6231560707092285, |
| "learning_rate": 0.0005615537118844952, |
| "loss": 3.9965, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.6565493488322032, |
| "grad_norm": 0.6274864077568054, |
| "learning_rate": 0.0005612304708544337, |
| "loss": 3.97, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.6619309008718114, |
| "grad_norm": 0.6003475189208984, |
| "learning_rate": 0.0005609072298243723, |
| "loss": 3.9817, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.6673124529114196, |
| "grad_norm": 0.6099907159805298, |
| "learning_rate": 0.0005605839887943109, |
| "loss": 3.9868, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.6726940049510278, |
| "grad_norm": 0.640876829624176, |
| "learning_rate": 0.0005602607477642495, |
| "loss": 3.9754, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.6780755569906362, |
| "grad_norm": 0.8232746720314026, |
| "learning_rate": 0.0005599375067341881, |
| "loss": 3.9749, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.6834571090302444, |
| "grad_norm": 0.6095024347305298, |
| "learning_rate": 0.0005596142657041266, |
| "loss": 3.9693, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.6888386610698526, |
| "grad_norm": 0.5760998725891113, |
| "learning_rate": 0.0005592910246740653, |
| "loss": 3.9533, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.6942202131094608, |
| "grad_norm": 0.6642665266990662, |
| "learning_rate": 0.0005589677836440038, |
| "loss": 3.9595, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.699601765149069, |
| "grad_norm": 0.5661361813545227, |
| "learning_rate": 0.0005586445426139425, |
| "loss": 3.9461, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.7049833171886772, |
| "grad_norm": 0.5967432856559753, |
| "learning_rate": 0.000558321301583881, |
| "loss": 3.9567, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.7103648692282855, |
| "grad_norm": 0.5816430449485779, |
| "learning_rate": 0.0005579980605538196, |
| "loss": 3.9586, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.7157464212678937, |
| "grad_norm": 0.5664747357368469, |
| "learning_rate": 0.0005576748195237582, |
| "loss": 3.9593, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.7211279733075019, |
| "grad_norm": 0.5324902534484863, |
| "learning_rate": 0.0005573515784936967, |
| "loss": 3.9369, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.7265095253471101, |
| "grad_norm": 0.8317891955375671, |
| "learning_rate": 0.0005570283374636353, |
| "loss": 3.9518, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.7318910773867183, |
| "grad_norm": 0.5899968147277832, |
| "learning_rate": 0.0005567050964335739, |
| "loss": 3.9414, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.7372726294263265, |
| "grad_norm": 0.5521121621131897, |
| "learning_rate": 0.0005563818554035126, |
| "loss": 3.9377, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.7426541814659348, |
| "grad_norm": 0.5316669940948486, |
| "learning_rate": 0.0005560586143734511, |
| "loss": 3.9372, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.748035733505543, |
| "grad_norm": 0.5491471290588379, |
| "learning_rate": 0.0005557353733433896, |
| "loss": 3.9333, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.7534172855451512, |
| "grad_norm": 0.5990865230560303, |
| "learning_rate": 0.0005554121323133283, |
| "loss": 3.937, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7534172855451512, |
| "eval_accuracy": 0.33351840541891065, |
| "eval_loss": 3.861109733581543, |
| "eval_runtime": 180.3493, |
| "eval_samples_per_second": 99.867, |
| "eval_steps_per_second": 6.243, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7587988375847594, |
| "grad_norm": 0.5951315760612488, |
| "learning_rate": 0.0005550888912832668, |
| "loss": 3.9214, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.7641803896243676, |
| "grad_norm": 0.6178938746452332, |
| "learning_rate": 0.0005547656502532055, |
| "loss": 3.9212, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.7695619416639758, |
| "grad_norm": 0.6568113565444946, |
| "learning_rate": 0.000554442409223144, |
| "loss": 3.917, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.7749434937035841, |
| "grad_norm": 0.6213955283164978, |
| "learning_rate": 0.0005541191681930826, |
| "loss": 3.9392, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.7803250457431924, |
| "grad_norm": 0.6062942743301392, |
| "learning_rate": 0.0005537959271630212, |
| "loss": 3.9189, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.7857065977828006, |
| "grad_norm": 0.6278512477874756, |
| "learning_rate": 0.0005534726861329598, |
| "loss": 3.908, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.7910881498224088, |
| "grad_norm": 0.5704114437103271, |
| "learning_rate": 0.0005531494451028983, |
| "loss": 3.9054, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.796469701862017, |
| "grad_norm": 0.579227089881897, |
| "learning_rate": 0.0005528262040728369, |
| "loss": 3.904, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.8018512539016253, |
| "grad_norm": 0.567130982875824, |
| "learning_rate": 0.0005525029630427755, |
| "loss": 3.8941, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.8072328059412335, |
| "grad_norm": 0.5331254005432129, |
| "learning_rate": 0.0005521797220127141, |
| "loss": 3.9149, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.8126143579808417, |
| "grad_norm": 0.5759610533714294, |
| "learning_rate": 0.0005518564809826527, |
| "loss": 3.9028, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.8179959100204499, |
| "grad_norm": 0.5558604001998901, |
| "learning_rate": 0.0005515332399525912, |
| "loss": 3.9277, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.8233774620600581, |
| "grad_norm": 0.6097841858863831, |
| "learning_rate": 0.0005512099989225299, |
| "loss": 3.8995, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.8287590140996663, |
| "grad_norm": 0.580385684967041, |
| "learning_rate": 0.0005508867578924685, |
| "loss": 3.8788, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.8341405661392746, |
| "grad_norm": 0.5963330268859863, |
| "learning_rate": 0.000550563516862407, |
| "loss": 3.87, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.8395221181788828, |
| "grad_norm": 0.6174511313438416, |
| "learning_rate": 0.0005502402758323456, |
| "loss": 3.9082, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.844903670218491, |
| "grad_norm": 0.5597873330116272, |
| "learning_rate": 0.0005499170348022841, |
| "loss": 3.8767, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.8502852222580992, |
| "grad_norm": 0.6453114151954651, |
| "learning_rate": 0.0005495937937722228, |
| "loss": 3.8749, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.8556667742977074, |
| "grad_norm": 0.6109079718589783, |
| "learning_rate": 0.0005492705527421614, |
| "loss": 3.8862, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.8610483263373157, |
| "grad_norm": 0.5893735885620117, |
| "learning_rate": 0.0005489473117121, |
| "loss": 3.8876, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8610483263373157, |
| "eval_accuracy": 0.33768654750124977, |
| "eval_loss": 3.8177008628845215, |
| "eval_runtime": 180.2298, |
| "eval_samples_per_second": 99.934, |
| "eval_steps_per_second": 6.248, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8664298783769239, |
| "grad_norm": 0.5334535241127014, |
| "learning_rate": 0.0005486240706820385, |
| "loss": 3.8913, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.8718114304165321, |
| "grad_norm": 0.6027869582176208, |
| "learning_rate": 0.0005483008296519772, |
| "loss": 3.8906, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.8771929824561403, |
| "grad_norm": 0.5669072270393372, |
| "learning_rate": 0.0005479775886219157, |
| "loss": 3.8675, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.8825745344957485, |
| "grad_norm": 0.5210901498794556, |
| "learning_rate": 0.0005476543475918542, |
| "loss": 3.8678, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.8879560865353568, |
| "grad_norm": 0.5795321464538574, |
| "learning_rate": 0.0005473311065617929, |
| "loss": 3.8764, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.8933376385749651, |
| "grad_norm": 0.5359807014465332, |
| "learning_rate": 0.0005470078655317314, |
| "loss": 3.87, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.8987191906145733, |
| "grad_norm": 0.5787212252616882, |
| "learning_rate": 0.0005466846245016701, |
| "loss": 3.8619, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.9041007426541815, |
| "grad_norm": 0.542587161064148, |
| "learning_rate": 0.0005463613834716086, |
| "loss": 3.8606, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.9094822946937897, |
| "grad_norm": 0.5083078742027283, |
| "learning_rate": 0.0005460381424415472, |
| "loss": 3.8736, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.9148638467333979, |
| "grad_norm": 0.5857490301132202, |
| "learning_rate": 0.0005457149014114858, |
| "loss": 3.8588, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.9202453987730062, |
| "grad_norm": 0.6556069850921631, |
| "learning_rate": 0.0005453916603814243, |
| "loss": 3.8751, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.9256269508126144, |
| "grad_norm": 0.6806515455245972, |
| "learning_rate": 0.000545068419351363, |
| "loss": 3.8549, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.9310085028522226, |
| "grad_norm": 0.5887377858161926, |
| "learning_rate": 0.0005447451783213015, |
| "loss": 3.8544, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.9363900548918308, |
| "grad_norm": 0.5550622344017029, |
| "learning_rate": 0.0005444219372912401, |
| "loss": 3.866, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.941771606931439, |
| "grad_norm": 0.5800760388374329, |
| "learning_rate": 0.0005440986962611787, |
| "loss": 3.837, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.9471531589710472, |
| "grad_norm": 0.6137863993644714, |
| "learning_rate": 0.0005437754552311173, |
| "loss": 3.8428, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.9525347110106555, |
| "grad_norm": 0.5557230710983276, |
| "learning_rate": 0.0005434522142010559, |
| "loss": 3.843, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.9579162630502637, |
| "grad_norm": 0.5875601172447205, |
| "learning_rate": 0.0005431289731709945, |
| "loss": 3.8428, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.9632978150898719, |
| "grad_norm": 0.5003922581672668, |
| "learning_rate": 0.000542805732140933, |
| "loss": 3.8504, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.9686793671294801, |
| "grad_norm": 0.6187765002250671, |
| "learning_rate": 0.0005424824911108716, |
| "loss": 3.8255, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9686793671294801, |
| "eval_accuracy": 0.3412266758920212, |
| "eval_loss": 3.7763822078704834, |
| "eval_runtime": 180.2292, |
| "eval_samples_per_second": 99.934, |
| "eval_steps_per_second": 6.248, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9740609191690883, |
| "grad_norm": 0.5244655609130859, |
| "learning_rate": 0.0005421592500808102, |
| "loss": 3.8325, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.9794424712086965, |
| "grad_norm": 0.6028016209602356, |
| "learning_rate": 0.0005418360090507488, |
| "loss": 3.8239, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.9848240232483048, |
| "grad_norm": 0.6184064745903015, |
| "learning_rate": 0.0005415127680206874, |
| "loss": 3.8494, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.9902055752879131, |
| "grad_norm": 0.5826838612556458, |
| "learning_rate": 0.0005411895269906259, |
| "loss": 3.8364, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.9955871273275213, |
| "grad_norm": 0.5381140112876892, |
| "learning_rate": 0.0005408662859605646, |
| "loss": 3.8343, |
| "step": 9250 |
| }, |
| { |
| "epoch": 1.0009686793671295, |
| "grad_norm": 0.7138656377792358, |
| "learning_rate": 0.0005405430449305031, |
| "loss": 3.8177, |
| "step": 9300 |
| }, |
| { |
| "epoch": 1.0063502314067376, |
| "grad_norm": 0.6079965233802795, |
| "learning_rate": 0.0005402198039004416, |
| "loss": 3.7709, |
| "step": 9350 |
| }, |
| { |
| "epoch": 1.011731783446346, |
| "grad_norm": 0.48963576555252075, |
| "learning_rate": 0.0005398965628703803, |
| "loss": 3.7634, |
| "step": 9400 |
| }, |
| { |
| "epoch": 1.017113335485954, |
| "grad_norm": 0.5836416482925415, |
| "learning_rate": 0.0005395733218403189, |
| "loss": 3.7652, |
| "step": 9450 |
| }, |
| { |
| "epoch": 1.0224948875255624, |
| "grad_norm": 0.6434711217880249, |
| "learning_rate": 0.0005392500808102575, |
| "loss": 3.777, |
| "step": 9500 |
| }, |
| { |
| "epoch": 1.0278764395651705, |
| "grad_norm": 0.5500202178955078, |
| "learning_rate": 0.000538926839780196, |
| "loss": 3.7877, |
| "step": 9550 |
| }, |
| { |
| "epoch": 1.0332579916047788, |
| "grad_norm": 0.5199800133705139, |
| "learning_rate": 0.0005386035987501346, |
| "loss": 3.774, |
| "step": 9600 |
| }, |
| { |
| "epoch": 1.0386395436443872, |
| "grad_norm": 0.5320031046867371, |
| "learning_rate": 0.0005382803577200732, |
| "loss": 3.78, |
| "step": 9650 |
| }, |
| { |
| "epoch": 1.0440210956839953, |
| "grad_norm": 0.5856112837791443, |
| "learning_rate": 0.0005379571166900119, |
| "loss": 3.7726, |
| "step": 9700 |
| }, |
| { |
| "epoch": 1.0494026477236036, |
| "grad_norm": 0.595620334148407, |
| "learning_rate": 0.0005376338756599504, |
| "loss": 3.7815, |
| "step": 9750 |
| }, |
| { |
| "epoch": 1.0547841997632117, |
| "grad_norm": 0.5389620661735535, |
| "learning_rate": 0.0005373106346298889, |
| "loss": 3.7741, |
| "step": 9800 |
| }, |
| { |
| "epoch": 1.06016575180282, |
| "grad_norm": 0.6535997986793518, |
| "learning_rate": 0.0005369873935998276, |
| "loss": 3.7702, |
| "step": 9850 |
| }, |
| { |
| "epoch": 1.0655473038424281, |
| "grad_norm": 0.6863070726394653, |
| "learning_rate": 0.0005366641525697661, |
| "loss": 3.7684, |
| "step": 9900 |
| }, |
| { |
| "epoch": 1.0709288558820365, |
| "grad_norm": 0.5795710682868958, |
| "learning_rate": 0.0005363409115397048, |
| "loss": 3.7652, |
| "step": 9950 |
| }, |
| { |
| "epoch": 1.0763104079216446, |
| "grad_norm": 0.6301584243774414, |
| "learning_rate": 0.0005360176705096433, |
| "loss": 3.7842, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0763104079216446, |
| "eval_accuracy": 0.34461621138028176, |
| "eval_loss": 3.7496469020843506, |
| "eval_runtime": 180.3887, |
| "eval_samples_per_second": 99.845, |
| "eval_steps_per_second": 6.242, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.081691959961253, |
| "grad_norm": 0.6134398579597473, |
| "learning_rate": 0.0005356944294795819, |
| "loss": 3.7578, |
| "step": 10050 |
| }, |
| { |
| "epoch": 1.087073512000861, |
| "grad_norm": 0.5343470573425293, |
| "learning_rate": 0.0005353711884495205, |
| "loss": 3.7574, |
| "step": 10100 |
| }, |
| { |
| "epoch": 1.0924550640404693, |
| "grad_norm": 0.5602631568908691, |
| "learning_rate": 0.000535047947419459, |
| "loss": 3.7551, |
| "step": 10150 |
| }, |
| { |
| "epoch": 1.0978366160800774, |
| "grad_norm": 0.56365966796875, |
| "learning_rate": 0.0005347247063893976, |
| "loss": 3.7695, |
| "step": 10200 |
| }, |
| { |
| "epoch": 1.1032181681196858, |
| "grad_norm": 0.5774141550064087, |
| "learning_rate": 0.0005344014653593362, |
| "loss": 3.7648, |
| "step": 10250 |
| }, |
| { |
| "epoch": 1.1085997201592939, |
| "grad_norm": 0.5826075077056885, |
| "learning_rate": 0.0005340782243292748, |
| "loss": 3.758, |
| "step": 10300 |
| }, |
| { |
| "epoch": 1.1139812721989022, |
| "grad_norm": 0.5574244856834412, |
| "learning_rate": 0.0005337549832992134, |
| "loss": 3.7457, |
| "step": 10350 |
| }, |
| { |
| "epoch": 1.1193628242385103, |
| "grad_norm": 0.5108634829521179, |
| "learning_rate": 0.000533431742269152, |
| "loss": 3.7676, |
| "step": 10400 |
| }, |
| { |
| "epoch": 1.1247443762781186, |
| "grad_norm": 0.5607388019561768, |
| "learning_rate": 0.0005331085012390905, |
| "loss": 3.7618, |
| "step": 10450 |
| }, |
| { |
| "epoch": 1.1301259283177267, |
| "grad_norm": 0.5880976319313049, |
| "learning_rate": 0.0005327852602090292, |
| "loss": 3.7593, |
| "step": 10500 |
| }, |
| { |
| "epoch": 1.135507480357335, |
| "grad_norm": 0.5758063793182373, |
| "learning_rate": 0.0005324620191789678, |
| "loss": 3.7638, |
| "step": 10550 |
| }, |
| { |
| "epoch": 1.1408890323969434, |
| "grad_norm": 0.5253919959068298, |
| "learning_rate": 0.0005321387781489063, |
| "loss": 3.7666, |
| "step": 10600 |
| }, |
| { |
| "epoch": 1.1462705844365515, |
| "grad_norm": 0.5626659393310547, |
| "learning_rate": 0.0005318155371188449, |
| "loss": 3.734, |
| "step": 10650 |
| }, |
| { |
| "epoch": 1.1516521364761596, |
| "grad_norm": 0.5251621603965759, |
| "learning_rate": 0.0005314922960887834, |
| "loss": 3.7614, |
| "step": 10700 |
| }, |
| { |
| "epoch": 1.157033688515768, |
| "grad_norm": 0.6023606061935425, |
| "learning_rate": 0.0005311690550587221, |
| "loss": 3.768, |
| "step": 10750 |
| }, |
| { |
| "epoch": 1.1624152405553763, |
| "grad_norm": 0.5632306933403015, |
| "learning_rate": 0.0005308458140286607, |
| "loss": 3.7378, |
| "step": 10800 |
| }, |
| { |
| "epoch": 1.1677967925949844, |
| "grad_norm": 0.5231441855430603, |
| "learning_rate": 0.0005305225729985993, |
| "loss": 3.7404, |
| "step": 10850 |
| }, |
| { |
| "epoch": 1.1731783446345927, |
| "grad_norm": 0.567490816116333, |
| "learning_rate": 0.0005301993319685378, |
| "loss": 3.7422, |
| "step": 10900 |
| }, |
| { |
| "epoch": 1.1785598966742008, |
| "grad_norm": 0.5332131385803223, |
| "learning_rate": 0.0005298760909384765, |
| "loss": 3.7561, |
| "step": 10950 |
| }, |
| { |
| "epoch": 1.1839414487138091, |
| "grad_norm": 0.5323898792266846, |
| "learning_rate": 0.000529552849908415, |
| "loss": 3.7212, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1839414487138091, |
| "eval_accuracy": 0.34692530255211507, |
| "eval_loss": 3.722130298614502, |
| "eval_runtime": 180.2165, |
| "eval_samples_per_second": 99.941, |
| "eval_steps_per_second": 6.248, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1893230007534172, |
| "grad_norm": 0.5329837203025818, |
| "learning_rate": 0.0005292296088783535, |
| "loss": 3.7392, |
| "step": 11050 |
| }, |
| { |
| "epoch": 1.1947045527930256, |
| "grad_norm": 0.6097086071968079, |
| "learning_rate": 0.0005289063678482922, |
| "loss": 3.7493, |
| "step": 11100 |
| }, |
| { |
| "epoch": 1.2000861048326337, |
| "grad_norm": 0.5243571996688843, |
| "learning_rate": 0.0005285831268182307, |
| "loss": 3.7329, |
| "step": 11150 |
| }, |
| { |
| "epoch": 1.205467656872242, |
| "grad_norm": 0.5530942678451538, |
| "learning_rate": 0.0005282598857881694, |
| "loss": 3.7355, |
| "step": 11200 |
| }, |
| { |
| "epoch": 1.21084920891185, |
| "grad_norm": 0.5588470101356506, |
| "learning_rate": 0.0005279366447581079, |
| "loss": 3.7601, |
| "step": 11250 |
| }, |
| { |
| "epoch": 1.2162307609514584, |
| "grad_norm": 0.5571747422218323, |
| "learning_rate": 0.0005276134037280465, |
| "loss": 3.7339, |
| "step": 11300 |
| }, |
| { |
| "epoch": 1.2216123129910665, |
| "grad_norm": 0.5556563138961792, |
| "learning_rate": 0.0005272901626979851, |
| "loss": 3.7238, |
| "step": 11350 |
| }, |
| { |
| "epoch": 1.2269938650306749, |
| "grad_norm": 0.6499513387680054, |
| "learning_rate": 0.0005269669216679236, |
| "loss": 3.7475, |
| "step": 11400 |
| }, |
| { |
| "epoch": 1.232375417070283, |
| "grad_norm": 0.5881815552711487, |
| "learning_rate": 0.0005266436806378623, |
| "loss": 3.7419, |
| "step": 11450 |
| }, |
| { |
| "epoch": 1.2377569691098913, |
| "grad_norm": 0.5554115772247314, |
| "learning_rate": 0.0005263204396078008, |
| "loss": 3.7435, |
| "step": 11500 |
| }, |
| { |
| "epoch": 1.2431385211494996, |
| "grad_norm": 0.562907338142395, |
| "learning_rate": 0.0005259971985777394, |
| "loss": 3.7435, |
| "step": 11550 |
| }, |
| { |
| "epoch": 1.2485200731891077, |
| "grad_norm": 0.5462493896484375, |
| "learning_rate": 0.000525673957547678, |
| "loss": 3.7375, |
| "step": 11600 |
| }, |
| { |
| "epoch": 1.2539016252287158, |
| "grad_norm": 0.5715740323066711, |
| "learning_rate": 0.0005253507165176167, |
| "loss": 3.732, |
| "step": 11650 |
| }, |
| { |
| "epoch": 1.2592831772683242, |
| "grad_norm": 0.5162888169288635, |
| "learning_rate": 0.0005250274754875552, |
| "loss": 3.7298, |
| "step": 11700 |
| }, |
| { |
| "epoch": 1.2646647293079325, |
| "grad_norm": 0.5549124479293823, |
| "learning_rate": 0.0005247042344574938, |
| "loss": 3.73, |
| "step": 11750 |
| }, |
| { |
| "epoch": 1.2700462813475406, |
| "grad_norm": 0.5421045422554016, |
| "learning_rate": 0.0005243809934274323, |
| "loss": 3.7321, |
| "step": 11800 |
| }, |
| { |
| "epoch": 1.275427833387149, |
| "grad_norm": 0.5627176761627197, |
| "learning_rate": 0.0005240577523973709, |
| "loss": 3.7224, |
| "step": 11850 |
| }, |
| { |
| "epoch": 1.280809385426757, |
| "grad_norm": 0.5783089995384216, |
| "learning_rate": 0.0005237345113673095, |
| "loss": 3.7265, |
| "step": 11900 |
| }, |
| { |
| "epoch": 1.2861909374663654, |
| "grad_norm": 0.6225681900978088, |
| "learning_rate": 0.0005234112703372481, |
| "loss": 3.7381, |
| "step": 11950 |
| }, |
| { |
| "epoch": 1.2915724895059735, |
| "grad_norm": 0.5271321535110474, |
| "learning_rate": 0.0005230880293071867, |
| "loss": 3.7296, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2915724895059735, |
| "eval_accuracy": 0.3496377132435158, |
| "eval_loss": 3.698834180831909, |
| "eval_runtime": 179.7345, |
| "eval_samples_per_second": 100.209, |
| "eval_steps_per_second": 6.265, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2969540415455818, |
| "grad_norm": 0.5290063619613647, |
| "learning_rate": 0.0005227647882771253, |
| "loss": 3.7369, |
| "step": 12050 |
| }, |
| { |
| "epoch": 1.30233559358519, |
| "grad_norm": 0.5177344679832458, |
| "learning_rate": 0.0005224415472470639, |
| "loss": 3.7374, |
| "step": 12100 |
| }, |
| { |
| "epoch": 1.3077171456247982, |
| "grad_norm": 0.5714365839958191, |
| "learning_rate": 0.0005221247710376037, |
| "loss": 3.7489, |
| "step": 12150 |
| }, |
| { |
| "epoch": 1.3130986976644063, |
| "grad_norm": 0.5795943140983582, |
| "learning_rate": 0.0005218015300075422, |
| "loss": 3.7384, |
| "step": 12200 |
| }, |
| { |
| "epoch": 1.3184802497040147, |
| "grad_norm": 0.5433863401412964, |
| "learning_rate": 0.0005214782889774809, |
| "loss": 3.7228, |
| "step": 12250 |
| }, |
| { |
| "epoch": 1.3238618017436228, |
| "grad_norm": 0.5290769338607788, |
| "learning_rate": 0.0005211550479474194, |
| "loss": 3.7153, |
| "step": 12300 |
| }, |
| { |
| "epoch": 1.329243353783231, |
| "grad_norm": 0.6278223395347595, |
| "learning_rate": 0.000520831806917358, |
| "loss": 3.7361, |
| "step": 12350 |
| }, |
| { |
| "epoch": 1.3346249058228392, |
| "grad_norm": 0.5160257816314697, |
| "learning_rate": 0.0005205085658872966, |
| "loss": 3.6936, |
| "step": 12400 |
| }, |
| { |
| "epoch": 1.3400064578624475, |
| "grad_norm": 0.591951310634613, |
| "learning_rate": 0.0005201853248572351, |
| "loss": 3.7249, |
| "step": 12450 |
| }, |
| { |
| "epoch": 1.3453880099020559, |
| "grad_norm": 0.5568095445632935, |
| "learning_rate": 0.0005198620838271738, |
| "loss": 3.726, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.350769561941664, |
| "grad_norm": 0.5874963998794556, |
| "learning_rate": 0.0005195388427971123, |
| "loss": 3.7238, |
| "step": 12550 |
| }, |
| { |
| "epoch": 1.356151113981272, |
| "grad_norm": 0.5767421126365662, |
| "learning_rate": 0.000519215601767051, |
| "loss": 3.7266, |
| "step": 12600 |
| }, |
| { |
| "epoch": 1.3615326660208804, |
| "grad_norm": 0.5423089265823364, |
| "learning_rate": 0.0005188923607369895, |
| "loss": 3.7133, |
| "step": 12650 |
| }, |
| { |
| "epoch": 1.3669142180604887, |
| "grad_norm": 0.562261164188385, |
| "learning_rate": 0.000518569119706928, |
| "loss": 3.7204, |
| "step": 12700 |
| }, |
| { |
| "epoch": 1.3722957701000968, |
| "grad_norm": 0.5939424633979797, |
| "learning_rate": 0.0005182458786768667, |
| "loss": 3.7343, |
| "step": 12750 |
| }, |
| { |
| "epoch": 1.3776773221397052, |
| "grad_norm": 0.5418099164962769, |
| "learning_rate": 0.0005179226376468052, |
| "loss": 3.7304, |
| "step": 12800 |
| }, |
| { |
| "epoch": 1.3830588741793133, |
| "grad_norm": 0.5787643790245056, |
| "learning_rate": 0.0005175993966167438, |
| "loss": 3.7152, |
| "step": 12850 |
| }, |
| { |
| "epoch": 1.3884404262189216, |
| "grad_norm": 0.5920687913894653, |
| "learning_rate": 0.0005172761555866824, |
| "loss": 3.7025, |
| "step": 12900 |
| }, |
| { |
| "epoch": 1.3938219782585297, |
| "grad_norm": 0.5622273683547974, |
| "learning_rate": 0.0005169529145566211, |
| "loss": 3.7297, |
| "step": 12950 |
| }, |
| { |
| "epoch": 1.399203530298138, |
| "grad_norm": 0.5593236684799194, |
| "learning_rate": 0.0005166296735265596, |
| "loss": 3.6772, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.399203530298138, |
| "eval_accuracy": 0.351960929290765, |
| "eval_loss": 3.674530506134033, |
| "eval_runtime": 180.1009, |
| "eval_samples_per_second": 100.005, |
| "eval_steps_per_second": 6.252, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4045850823377461, |
| "grad_norm": 0.551760733127594, |
| "learning_rate": 0.0005163064324964982, |
| "loss": 3.7236, |
| "step": 13050 |
| }, |
| { |
| "epoch": 1.4099666343773545, |
| "grad_norm": 0.48447367548942566, |
| "learning_rate": 0.0005159831914664367, |
| "loss": 3.7066, |
| "step": 13100 |
| }, |
| { |
| "epoch": 1.4153481864169626, |
| "grad_norm": 0.5610434412956238, |
| "learning_rate": 0.0005156599504363753, |
| "loss": 3.7186, |
| "step": 13150 |
| }, |
| { |
| "epoch": 1.420729738456571, |
| "grad_norm": 0.6366670727729797, |
| "learning_rate": 0.000515336709406314, |
| "loss": 3.7192, |
| "step": 13200 |
| }, |
| { |
| "epoch": 1.426111290496179, |
| "grad_norm": 0.6398788094520569, |
| "learning_rate": 0.0005150134683762525, |
| "loss": 3.7288, |
| "step": 13250 |
| }, |
| { |
| "epoch": 1.4314928425357873, |
| "grad_norm": 0.5352592468261719, |
| "learning_rate": 0.0005146902273461911, |
| "loss": 3.7099, |
| "step": 13300 |
| }, |
| { |
| "epoch": 1.4368743945753955, |
| "grad_norm": 0.5241870284080505, |
| "learning_rate": 0.0005143669863161297, |
| "loss": 3.7291, |
| "step": 13350 |
| }, |
| { |
| "epoch": 1.4422559466150038, |
| "grad_norm": 0.5873683094978333, |
| "learning_rate": 0.0005140437452860683, |
| "loss": 3.7056, |
| "step": 13400 |
| }, |
| { |
| "epoch": 1.447637498654612, |
| "grad_norm": 0.5123621225357056, |
| "learning_rate": 0.0005137205042560069, |
| "loss": 3.6975, |
| "step": 13450 |
| }, |
| { |
| "epoch": 1.4530190506942202, |
| "grad_norm": 0.5494991540908813, |
| "learning_rate": 0.0005133972632259455, |
| "loss": 3.7057, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.4584006027338283, |
| "grad_norm": 0.5739464163780212, |
| "learning_rate": 0.0005130804870164853, |
| "loss": 3.6991, |
| "step": 13550 |
| }, |
| { |
| "epoch": 1.4637821547734367, |
| "grad_norm": 0.5516660809516907, |
| "learning_rate": 0.0005127572459864238, |
| "loss": 3.7064, |
| "step": 13600 |
| }, |
| { |
| "epoch": 1.469163706813045, |
| "grad_norm": 0.5697253346443176, |
| "learning_rate": 0.0005124340049563624, |
| "loss": 3.6926, |
| "step": 13650 |
| }, |
| { |
| "epoch": 1.474545258852653, |
| "grad_norm": 0.5626788139343262, |
| "learning_rate": 0.000512110763926301, |
| "loss": 3.677, |
| "step": 13700 |
| }, |
| { |
| "epoch": 1.4799268108922612, |
| "grad_norm": 0.5842751860618591, |
| "learning_rate": 0.0005117875228962395, |
| "loss": 3.6764, |
| "step": 13750 |
| }, |
| { |
| "epoch": 1.4853083629318695, |
| "grad_norm": 0.5150489211082458, |
| "learning_rate": 0.0005114642818661782, |
| "loss": 3.6995, |
| "step": 13800 |
| }, |
| { |
| "epoch": 1.4906899149714778, |
| "grad_norm": 0.5614669322967529, |
| "learning_rate": 0.0005111410408361168, |
| "loss": 3.6865, |
| "step": 13850 |
| }, |
| { |
| "epoch": 1.496071467011086, |
| "grad_norm": 0.5543150901794434, |
| "learning_rate": 0.0005108177998060554, |
| "loss": 3.6665, |
| "step": 13900 |
| }, |
| { |
| "epoch": 1.501453019050694, |
| "grad_norm": 0.6134596467018127, |
| "learning_rate": 0.0005104945587759939, |
| "loss": 3.7014, |
| "step": 13950 |
| }, |
| { |
| "epoch": 1.5068345710903024, |
| "grad_norm": 0.564272403717041, |
| "learning_rate": 0.0005101713177459324, |
| "loss": 3.6926, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5068345710903024, |
| "eval_accuracy": 0.353645918274992, |
| "eval_loss": 3.6572258472442627, |
| "eval_runtime": 179.942, |
| "eval_samples_per_second": 100.093, |
| "eval_steps_per_second": 6.258, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5122161231299107, |
| "grad_norm": 0.5460367202758789, |
| "learning_rate": 0.0005098480767158711, |
| "loss": 3.7094, |
| "step": 14050 |
| }, |
| { |
| "epoch": 1.5175976751695188, |
| "grad_norm": 0.5200446844100952, |
| "learning_rate": 0.0005095248356858097, |
| "loss": 3.6915, |
| "step": 14100 |
| }, |
| { |
| "epoch": 1.5229792272091272, |
| "grad_norm": 0.5343534350395203, |
| "learning_rate": 0.0005092015946557483, |
| "loss": 3.6877, |
| "step": 14150 |
| }, |
| { |
| "epoch": 1.5283607792487355, |
| "grad_norm": 0.5287026166915894, |
| "learning_rate": 0.0005088783536256868, |
| "loss": 3.6859, |
| "step": 14200 |
| }, |
| { |
| "epoch": 1.5337423312883436, |
| "grad_norm": 0.5682664513587952, |
| "learning_rate": 0.0005085551125956255, |
| "loss": 3.6866, |
| "step": 14250 |
| }, |
| { |
| "epoch": 1.5391238833279517, |
| "grad_norm": 0.5513389110565186, |
| "learning_rate": 0.000508231871565564, |
| "loss": 3.6758, |
| "step": 14300 |
| }, |
| { |
| "epoch": 1.54450543536756, |
| "grad_norm": 0.5954504013061523, |
| "learning_rate": 0.0005079086305355026, |
| "loss": 3.6807, |
| "step": 14350 |
| }, |
| { |
| "epoch": 1.5498869874071683, |
| "grad_norm": 0.534307062625885, |
| "learning_rate": 0.0005075853895054412, |
| "loss": 3.7011, |
| "step": 14400 |
| }, |
| { |
| "epoch": 1.5552685394467765, |
| "grad_norm": 0.5764222145080566, |
| "learning_rate": 0.0005072621484753797, |
| "loss": 3.6999, |
| "step": 14450 |
| }, |
| { |
| "epoch": 1.5606500914863846, |
| "grad_norm": 0.5494203567504883, |
| "learning_rate": 0.0005069389074453184, |
| "loss": 3.6871, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.566031643525993, |
| "grad_norm": 0.5286200046539307, |
| "learning_rate": 0.0005066156664152569, |
| "loss": 3.6961, |
| "step": 14550 |
| }, |
| { |
| "epoch": 1.5714131955656012, |
| "grad_norm": 0.5253137350082397, |
| "learning_rate": 0.0005062924253851955, |
| "loss": 3.6849, |
| "step": 14600 |
| }, |
| { |
| "epoch": 1.5767947476052093, |
| "grad_norm": 0.5719892978668213, |
| "learning_rate": 0.0005059691843551341, |
| "loss": 3.6662, |
| "step": 14650 |
| }, |
| { |
| "epoch": 1.5821762996448174, |
| "grad_norm": 0.6287044882774353, |
| "learning_rate": 0.0005056459433250727, |
| "loss": 3.6673, |
| "step": 14700 |
| }, |
| { |
| "epoch": 1.5875578516844258, |
| "grad_norm": 0.5336293578147888, |
| "learning_rate": 0.0005053227022950113, |
| "loss": 3.6737, |
| "step": 14750 |
| }, |
| { |
| "epoch": 1.592939403724034, |
| "grad_norm": 0.5435810089111328, |
| "learning_rate": 0.0005049994612649499, |
| "loss": 3.6794, |
| "step": 14800 |
| }, |
| { |
| "epoch": 1.5983209557636422, |
| "grad_norm": 0.6014664769172668, |
| "learning_rate": 0.0005046762202348884, |
| "loss": 3.6995, |
| "step": 14850 |
| }, |
| { |
| "epoch": 1.6037025078032503, |
| "grad_norm": 0.5819555521011353, |
| "learning_rate": 0.000504352979204827, |
| "loss": 3.6859, |
| "step": 14900 |
| }, |
| { |
| "epoch": 1.6090840598428586, |
| "grad_norm": 0.541706919670105, |
| "learning_rate": 0.0005040297381747656, |
| "loss": 3.6838, |
| "step": 14950 |
| }, |
| { |
| "epoch": 1.614465611882467, |
| "grad_norm": 0.6835187077522278, |
| "learning_rate": 0.0005037064971447042, |
| "loss": 3.6872, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.614465611882467, |
| "eval_accuracy": 0.35586591407881746, |
| "eval_loss": 3.635637044906616, |
| "eval_runtime": 179.85, |
| "eval_samples_per_second": 100.145, |
| "eval_steps_per_second": 6.261, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.619847163922075, |
| "grad_norm": 0.5497548580169678, |
| "learning_rate": 0.0005033832561146428, |
| "loss": 3.6633, |
| "step": 15050 |
| }, |
| { |
| "epoch": 1.6252287159616834, |
| "grad_norm": 0.5530732274055481, |
| "learning_rate": 0.0005030600150845813, |
| "loss": 3.6751, |
| "step": 15100 |
| }, |
| { |
| "epoch": 1.6306102680012917, |
| "grad_norm": 0.5972110629081726, |
| "learning_rate": 0.00050273677405452, |
| "loss": 3.6909, |
| "step": 15150 |
| }, |
| { |
| "epoch": 1.6359918200408998, |
| "grad_norm": 0.5513889193534851, |
| "learning_rate": 0.0005024135330244585, |
| "loss": 3.6774, |
| "step": 15200 |
| }, |
| { |
| "epoch": 1.641373372080508, |
| "grad_norm": 0.5464329719543457, |
| "learning_rate": 0.0005020902919943972, |
| "loss": 3.6562, |
| "step": 15250 |
| }, |
| { |
| "epoch": 1.6467549241201163, |
| "grad_norm": 0.621970534324646, |
| "learning_rate": 0.0005017670509643357, |
| "loss": 3.6929, |
| "step": 15300 |
| }, |
| { |
| "epoch": 1.6521364761597246, |
| "grad_norm": 0.5582807064056396, |
| "learning_rate": 0.0005014438099342743, |
| "loss": 3.6783, |
| "step": 15350 |
| }, |
| { |
| "epoch": 1.6575180281993327, |
| "grad_norm": 0.5590953230857849, |
| "learning_rate": 0.0005011205689042129, |
| "loss": 3.6581, |
| "step": 15400 |
| }, |
| { |
| "epoch": 1.6628995802389408, |
| "grad_norm": 0.5335003137588501, |
| "learning_rate": 0.0005007973278741514, |
| "loss": 3.6591, |
| "step": 15450 |
| }, |
| { |
| "epoch": 1.6682811322785491, |
| "grad_norm": 0.5382925868034363, |
| "learning_rate": 0.00050047408684409, |
| "loss": 3.6654, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.6736626843181575, |
| "grad_norm": 0.5382738709449768, |
| "learning_rate": 0.0005001508458140286, |
| "loss": 3.6541, |
| "step": 15550 |
| }, |
| { |
| "epoch": 1.6790442363577656, |
| "grad_norm": 0.5367087721824646, |
| "learning_rate": 0.0004998276047839673, |
| "loss": 3.6594, |
| "step": 15600 |
| }, |
| { |
| "epoch": 1.6844257883973737, |
| "grad_norm": 0.5150970816612244, |
| "learning_rate": 0.0004995043637539058, |
| "loss": 3.6849, |
| "step": 15650 |
| }, |
| { |
| "epoch": 1.689807340436982, |
| "grad_norm": 0.5762928128242493, |
| "learning_rate": 0.0004991811227238443, |
| "loss": 3.6711, |
| "step": 15700 |
| }, |
| { |
| "epoch": 1.6951888924765903, |
| "grad_norm": 0.48318615555763245, |
| "learning_rate": 0.0004988578816937829, |
| "loss": 3.6631, |
| "step": 15750 |
| }, |
| { |
| "epoch": 1.7005704445161984, |
| "grad_norm": 0.5551878213882446, |
| "learning_rate": 0.0004985346406637215, |
| "loss": 3.6544, |
| "step": 15800 |
| }, |
| { |
| "epoch": 1.7059519965558065, |
| "grad_norm": 0.5169930458068848, |
| "learning_rate": 0.0004982113996336602, |
| "loss": 3.6696, |
| "step": 15850 |
| }, |
| { |
| "epoch": 1.7113335485954149, |
| "grad_norm": 0.592219352722168, |
| "learning_rate": 0.0004978881586035987, |
| "loss": 3.6806, |
| "step": 15900 |
| }, |
| { |
| "epoch": 1.7167151006350232, |
| "grad_norm": 0.492858350276947, |
| "learning_rate": 0.0004975649175735373, |
| "loss": 3.6541, |
| "step": 15950 |
| }, |
| { |
| "epoch": 1.7220966526746313, |
| "grad_norm": 0.5765631198883057, |
| "learning_rate": 0.0004972416765434759, |
| "loss": 3.6757, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7220966526746313, |
| "eval_accuracy": 0.3572902447851775, |
| "eval_loss": 3.61974835395813, |
| "eval_runtime": 179.8687, |
| "eval_samples_per_second": 100.134, |
| "eval_steps_per_second": 6.26, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7274782047142396, |
| "grad_norm": 0.6189604997634888, |
| "learning_rate": 0.0004969184355134145, |
| "loss": 3.6559, |
| "step": 16050 |
| }, |
| { |
| "epoch": 1.732859756753848, |
| "grad_norm": 0.5432648658752441, |
| "learning_rate": 0.0004965951944833531, |
| "loss": 3.6684, |
| "step": 16100 |
| }, |
| { |
| "epoch": 1.738241308793456, |
| "grad_norm": 0.6157057881355286, |
| "learning_rate": 0.0004962719534532916, |
| "loss": 3.6691, |
| "step": 16150 |
| }, |
| { |
| "epoch": 1.7436228608330642, |
| "grad_norm": 0.6409887671470642, |
| "learning_rate": 0.0004959487124232302, |
| "loss": 3.6738, |
| "step": 16200 |
| }, |
| { |
| "epoch": 1.7490044128726725, |
| "grad_norm": 0.5156955122947693, |
| "learning_rate": 0.0004956254713931688, |
| "loss": 3.673, |
| "step": 16250 |
| }, |
| { |
| "epoch": 1.7543859649122808, |
| "grad_norm": 0.5291411280632019, |
| "learning_rate": 0.0004953022303631074, |
| "loss": 3.6519, |
| "step": 16300 |
| }, |
| { |
| "epoch": 1.759767516951889, |
| "grad_norm": 0.7061535120010376, |
| "learning_rate": 0.0004949789893330459, |
| "loss": 3.656, |
| "step": 16350 |
| }, |
| { |
| "epoch": 1.765149068991497, |
| "grad_norm": 0.5498104095458984, |
| "learning_rate": 0.0004946557483029846, |
| "loss": 3.6604, |
| "step": 16400 |
| }, |
| { |
| "epoch": 1.7705306210311054, |
| "grad_norm": 0.560353696346283, |
| "learning_rate": 0.0004943325072729231, |
| "loss": 3.6492, |
| "step": 16450 |
| }, |
| { |
| "epoch": 1.7759121730707137, |
| "grad_norm": 0.5988339185714722, |
| "learning_rate": 0.0004940092662428617, |
| "loss": 3.6731, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.7812937251103218, |
| "grad_norm": 0.5460607409477234, |
| "learning_rate": 0.0004936860252128003, |
| "loss": 3.6579, |
| "step": 16550 |
| }, |
| { |
| "epoch": 1.78667527714993, |
| "grad_norm": 0.5597132444381714, |
| "learning_rate": 0.0004933627841827388, |
| "loss": 3.6591, |
| "step": 16600 |
| }, |
| { |
| "epoch": 1.7920568291895382, |
| "grad_norm": 0.589875340461731, |
| "learning_rate": 0.0004930395431526775, |
| "loss": 3.6633, |
| "step": 16650 |
| }, |
| { |
| "epoch": 1.7974383812291466, |
| "grad_norm": 0.5566060543060303, |
| "learning_rate": 0.0004927163021226161, |
| "loss": 3.648, |
| "step": 16700 |
| }, |
| { |
| "epoch": 1.8028199332687547, |
| "grad_norm": 0.5171520709991455, |
| "learning_rate": 0.0004923930610925547, |
| "loss": 3.646, |
| "step": 16750 |
| }, |
| { |
| "epoch": 1.8082014853083628, |
| "grad_norm": 0.5987152457237244, |
| "learning_rate": 0.0004920698200624932, |
| "loss": 3.6482, |
| "step": 16800 |
| }, |
| { |
| "epoch": 1.813583037347971, |
| "grad_norm": 0.5647301077842712, |
| "learning_rate": 0.0004917465790324317, |
| "loss": 3.624, |
| "step": 16850 |
| }, |
| { |
| "epoch": 1.8189645893875794, |
| "grad_norm": 0.5267931818962097, |
| "learning_rate": 0.0004914233380023704, |
| "loss": 3.6343, |
| "step": 16900 |
| }, |
| { |
| "epoch": 1.8243461414271875, |
| "grad_norm": 0.5864246487617493, |
| "learning_rate": 0.0004911000969723089, |
| "loss": 3.6659, |
| "step": 16950 |
| }, |
| { |
| "epoch": 1.8297276934667959, |
| "grad_norm": 0.5181066393852234, |
| "learning_rate": 0.0004907768559422476, |
| "loss": 3.6485, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8297276934667959, |
| "eval_accuracy": 0.3588785326992496, |
| "eval_loss": 3.603902816772461, |
| "eval_runtime": 180.6484, |
| "eval_samples_per_second": 99.702, |
| "eval_steps_per_second": 6.233, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8351092455064042, |
| "grad_norm": 0.5298339128494263, |
| "learning_rate": 0.0004904536149121861, |
| "loss": 3.6723, |
| "step": 17050 |
| }, |
| { |
| "epoch": 1.8404907975460123, |
| "grad_norm": 0.5916500091552734, |
| "learning_rate": 0.0004901303738821248, |
| "loss": 3.643, |
| "step": 17100 |
| }, |
| { |
| "epoch": 1.8458723495856204, |
| "grad_norm": 0.6232360601425171, |
| "learning_rate": 0.0004898071328520633, |
| "loss": 3.6451, |
| "step": 17150 |
| }, |
| { |
| "epoch": 1.8512539016252287, |
| "grad_norm": 0.5699940323829651, |
| "learning_rate": 0.0004894838918220019, |
| "loss": 3.6625, |
| "step": 17200 |
| }, |
| { |
| "epoch": 1.856635453664837, |
| "grad_norm": 0.5958313345909119, |
| "learning_rate": 0.0004891606507919405, |
| "loss": 3.6632, |
| "step": 17250 |
| }, |
| { |
| "epoch": 1.8620170057044452, |
| "grad_norm": 0.5603935718536377, |
| "learning_rate": 0.000488837409761879, |
| "loss": 3.6399, |
| "step": 17300 |
| }, |
| { |
| "epoch": 1.8673985577440533, |
| "grad_norm": 0.5688992738723755, |
| "learning_rate": 0.0004885141687318177, |
| "loss": 3.6626, |
| "step": 17350 |
| }, |
| { |
| "epoch": 1.8727801097836616, |
| "grad_norm": 0.5361830592155457, |
| "learning_rate": 0.00048819092770175623, |
| "loss": 3.6463, |
| "step": 17400 |
| }, |
| { |
| "epoch": 1.87816166182327, |
| "grad_norm": 0.5704878568649292, |
| "learning_rate": 0.0004878676866716948, |
| "loss": 3.6309, |
| "step": 17450 |
| }, |
| { |
| "epoch": 1.883543213862878, |
| "grad_norm": 0.5955484509468079, |
| "learning_rate": 0.00048754444564163337, |
| "loss": 3.6267, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.8889247659024861, |
| "grad_norm": 0.5311874151229858, |
| "learning_rate": 0.000487221204611572, |
| "loss": 3.6326, |
| "step": 17550 |
| }, |
| { |
| "epoch": 1.8943063179420945, |
| "grad_norm": 0.5513229966163635, |
| "learning_rate": 0.00048689796358151056, |
| "loss": 3.6471, |
| "step": 17600 |
| }, |
| { |
| "epoch": 1.8996878699817028, |
| "grad_norm": 0.560566782951355, |
| "learning_rate": 0.00048657472255144915, |
| "loss": 3.6473, |
| "step": 17650 |
| }, |
| { |
| "epoch": 1.905069422021311, |
| "grad_norm": 0.5173212885856628, |
| "learning_rate": 0.00048625794634198896, |
| "loss": 3.6304, |
| "step": 17700 |
| }, |
| { |
| "epoch": 1.910450974060919, |
| "grad_norm": 0.550670862197876, |
| "learning_rate": 0.00048593470531192756, |
| "loss": 3.62, |
| "step": 17750 |
| }, |
| { |
| "epoch": 1.9158325261005273, |
| "grad_norm": 0.5653765797615051, |
| "learning_rate": 0.00048561146428186615, |
| "loss": 3.6364, |
| "step": 17800 |
| }, |
| { |
| "epoch": 1.9212140781401357, |
| "grad_norm": 0.4973013401031494, |
| "learning_rate": 0.0004852882232518047, |
| "loss": 3.623, |
| "step": 17850 |
| }, |
| { |
| "epoch": 1.9265956301797438, |
| "grad_norm": 0.5724383592605591, |
| "learning_rate": 0.00048496498222174334, |
| "loss": 3.6446, |
| "step": 17900 |
| }, |
| { |
| "epoch": 1.931977182219352, |
| "grad_norm": 0.49556764960289, |
| "learning_rate": 0.00048464174119168193, |
| "loss": 3.6491, |
| "step": 17950 |
| }, |
| { |
| "epoch": 1.9373587342589604, |
| "grad_norm": 0.5692288875579834, |
| "learning_rate": 0.0004843185001616205, |
| "loss": 3.6324, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9373587342589604, |
| "eval_accuracy": 0.3603584936841706, |
| "eval_loss": 3.5903141498565674, |
| "eval_runtime": 179.9609, |
| "eval_samples_per_second": 100.083, |
| "eval_steps_per_second": 6.257, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9427402862985685, |
| "grad_norm": 0.5478717088699341, |
| "learning_rate": 0.00048399525913155907, |
| "loss": 3.6503, |
| "step": 18050 |
| }, |
| { |
| "epoch": 1.9481218383381766, |
| "grad_norm": 0.5264731049537659, |
| "learning_rate": 0.0004836720181014976, |
| "loss": 3.6285, |
| "step": 18100 |
| }, |
| { |
| "epoch": 1.953503390377785, |
| "grad_norm": 0.5305705070495605, |
| "learning_rate": 0.0004833487770714362, |
| "loss": 3.6198, |
| "step": 18150 |
| }, |
| { |
| "epoch": 1.9588849424173933, |
| "grad_norm": 0.5922155976295471, |
| "learning_rate": 0.00048302553604137485, |
| "loss": 3.6275, |
| "step": 18200 |
| }, |
| { |
| "epoch": 1.9642664944570014, |
| "grad_norm": 0.5468623042106628, |
| "learning_rate": 0.0004827022950113134, |
| "loss": 3.6447, |
| "step": 18250 |
| }, |
| { |
| "epoch": 1.9696480464966095, |
| "grad_norm": 0.5142163038253784, |
| "learning_rate": 0.000482379053981252, |
| "loss": 3.6241, |
| "step": 18300 |
| }, |
| { |
| "epoch": 1.9750295985362178, |
| "grad_norm": 0.5266579389572144, |
| "learning_rate": 0.0004820558129511906, |
| "loss": 3.6267, |
| "step": 18350 |
| }, |
| { |
| "epoch": 1.9804111505758262, |
| "grad_norm": 0.5678869485855103, |
| "learning_rate": 0.0004817325719211291, |
| "loss": 3.6415, |
| "step": 18400 |
| }, |
| { |
| "epoch": 1.9857927026154343, |
| "grad_norm": 0.5284473299980164, |
| "learning_rate": 0.0004814093308910677, |
| "loss": 3.624, |
| "step": 18450 |
| }, |
| { |
| "epoch": 1.9911742546550424, |
| "grad_norm": 0.534938395023346, |
| "learning_rate": 0.00048108608986100637, |
| "loss": 3.6337, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.9965558066946507, |
| "grad_norm": 0.5706211924552917, |
| "learning_rate": 0.0004807628488309449, |
| "loss": 3.6429, |
| "step": 18550 |
| }, |
| { |
| "epoch": 2.001937358734259, |
| "grad_norm": 0.5797092914581299, |
| "learning_rate": 0.0004804396078008835, |
| "loss": 3.5995, |
| "step": 18600 |
| }, |
| { |
| "epoch": 2.007318910773867, |
| "grad_norm": 0.54485023021698, |
| "learning_rate": 0.00048011636677082204, |
| "loss": 3.5261, |
| "step": 18650 |
| }, |
| { |
| "epoch": 2.0127004628134753, |
| "grad_norm": 0.5473525524139404, |
| "learning_rate": 0.00047979312574076064, |
| "loss": 3.5536, |
| "step": 18700 |
| }, |
| { |
| "epoch": 2.018082014853084, |
| "grad_norm": 0.556251049041748, |
| "learning_rate": 0.0004794698847106992, |
| "loss": 3.5245, |
| "step": 18750 |
| }, |
| { |
| "epoch": 2.023463566892692, |
| "grad_norm": 0.5699254870414734, |
| "learning_rate": 0.0004791466436806378, |
| "loss": 3.5467, |
| "step": 18800 |
| }, |
| { |
| "epoch": 2.0288451189323, |
| "grad_norm": 0.5414472222328186, |
| "learning_rate": 0.0004788234026505764, |
| "loss": 3.5511, |
| "step": 18850 |
| }, |
| { |
| "epoch": 2.034226670971908, |
| "grad_norm": 0.5470781922340393, |
| "learning_rate": 0.00047850016162051496, |
| "loss": 3.5416, |
| "step": 18900 |
| }, |
| { |
| "epoch": 2.0396082230115167, |
| "grad_norm": 0.5483036041259766, |
| "learning_rate": 0.00047817692059045356, |
| "loss": 3.5396, |
| "step": 18950 |
| }, |
| { |
| "epoch": 2.044989775051125, |
| "grad_norm": 0.5391370058059692, |
| "learning_rate": 0.00047785367956039215, |
| "loss": 3.5384, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.044989775051125, |
| "eval_accuracy": 0.36190886174039544, |
| "eval_loss": 3.5802509784698486, |
| "eval_runtime": 179.9777, |
| "eval_samples_per_second": 100.074, |
| "eval_steps_per_second": 6.256, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.050371327090733, |
| "grad_norm": 0.6000381708145142, |
| "learning_rate": 0.00047753043853033075, |
| "loss": 3.5573, |
| "step": 19050 |
| }, |
| { |
| "epoch": 2.055752879130341, |
| "grad_norm": 0.5803197026252747, |
| "learning_rate": 0.00047720719750026934, |
| "loss": 3.5533, |
| "step": 19100 |
| }, |
| { |
| "epoch": 2.0611344311699495, |
| "grad_norm": 0.6103200912475586, |
| "learning_rate": 0.00047688395647020793, |
| "loss": 3.5525, |
| "step": 19150 |
| }, |
| { |
| "epoch": 2.0665159832095576, |
| "grad_norm": 0.5805879831314087, |
| "learning_rate": 0.0004765607154401465, |
| "loss": 3.5686, |
| "step": 19200 |
| }, |
| { |
| "epoch": 2.0718975352491658, |
| "grad_norm": 0.5827572345733643, |
| "learning_rate": 0.00047623747441008507, |
| "loss": 3.5494, |
| "step": 19250 |
| }, |
| { |
| "epoch": 2.0772790872887743, |
| "grad_norm": 0.6824135184288025, |
| "learning_rate": 0.0004759142333800236, |
| "loss": 3.551, |
| "step": 19300 |
| }, |
| { |
| "epoch": 2.0826606393283824, |
| "grad_norm": 0.5492932200431824, |
| "learning_rate": 0.00047559099234996226, |
| "loss": 3.5614, |
| "step": 19350 |
| }, |
| { |
| "epoch": 2.0880421913679905, |
| "grad_norm": 0.5775820016860962, |
| "learning_rate": 0.00047526775131990085, |
| "loss": 3.5566, |
| "step": 19400 |
| }, |
| { |
| "epoch": 2.0934237434075986, |
| "grad_norm": 0.5697767734527588, |
| "learning_rate": 0.0004749445102898394, |
| "loss": 3.5567, |
| "step": 19450 |
| }, |
| { |
| "epoch": 2.098805295447207, |
| "grad_norm": 0.5625877380371094, |
| "learning_rate": 0.000474621269259778, |
| "loss": 3.5373, |
| "step": 19500 |
| }, |
| { |
| "epoch": 2.1041868474868153, |
| "grad_norm": 0.5145209431648254, |
| "learning_rate": 0.0004742980282297166, |
| "loss": 3.5329, |
| "step": 19550 |
| }, |
| { |
| "epoch": 2.1095683995264234, |
| "grad_norm": 0.5731906294822693, |
| "learning_rate": 0.0004739747871996551, |
| "loss": 3.5426, |
| "step": 19600 |
| }, |
| { |
| "epoch": 2.1149499515660315, |
| "grad_norm": 0.5850881934165955, |
| "learning_rate": 0.00047365154616959377, |
| "loss": 3.536, |
| "step": 19650 |
| }, |
| { |
| "epoch": 2.12033150360564, |
| "grad_norm": 0.5699858069419861, |
| "learning_rate": 0.00047332830513953237, |
| "loss": 3.5535, |
| "step": 19700 |
| }, |
| { |
| "epoch": 2.125713055645248, |
| "grad_norm": 0.6195251941680908, |
| "learning_rate": 0.0004730050641094709, |
| "loss": 3.5599, |
| "step": 19750 |
| }, |
| { |
| "epoch": 2.1310946076848563, |
| "grad_norm": 0.598713755607605, |
| "learning_rate": 0.0004726818230794095, |
| "loss": 3.5721, |
| "step": 19800 |
| }, |
| { |
| "epoch": 2.1364761597244644, |
| "grad_norm": 0.5637555718421936, |
| "learning_rate": 0.0004723650468699493, |
| "loss": 3.5659, |
| "step": 19850 |
| }, |
| { |
| "epoch": 2.141857711764073, |
| "grad_norm": 0.6245604753494263, |
| "learning_rate": 0.0004720418058398879, |
| "loss": 3.56, |
| "step": 19900 |
| }, |
| { |
| "epoch": 2.147239263803681, |
| "grad_norm": 0.5333353877067566, |
| "learning_rate": 0.00047171856480982644, |
| "loss": 3.5668, |
| "step": 19950 |
| }, |
| { |
| "epoch": 2.152620815843289, |
| "grad_norm": 0.5466655492782593, |
| "learning_rate": 0.0004713953237797651, |
| "loss": 3.5558, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.152620815843289, |
| "eval_accuracy": 0.3628188296758417, |
| "eval_loss": 3.5698418617248535, |
| "eval_runtime": 180.3488, |
| "eval_samples_per_second": 99.868, |
| "eval_steps_per_second": 6.243, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.1580023678828972, |
| "grad_norm": 0.5573538541793823, |
| "learning_rate": 0.0004710720827497037, |
| "loss": 3.5583, |
| "step": 20050 |
| }, |
| { |
| "epoch": 2.163383919922506, |
| "grad_norm": 0.5291585326194763, |
| "learning_rate": 0.00047074884171964223, |
| "loss": 3.5514, |
| "step": 20100 |
| }, |
| { |
| "epoch": 2.168765471962114, |
| "grad_norm": 0.5464843511581421, |
| "learning_rate": 0.0004704256006895808, |
| "loss": 3.5752, |
| "step": 20150 |
| }, |
| { |
| "epoch": 2.174147024001722, |
| "grad_norm": 0.5665469765663147, |
| "learning_rate": 0.00047010235965951936, |
| "loss": 3.5353, |
| "step": 20200 |
| }, |
| { |
| "epoch": 2.1795285760413305, |
| "grad_norm": 0.5358191728591919, |
| "learning_rate": 0.00046977911862945796, |
| "loss": 3.5741, |
| "step": 20250 |
| }, |
| { |
| "epoch": 2.1849101280809387, |
| "grad_norm": 0.6494370698928833, |
| "learning_rate": 0.0004694558775993966, |
| "loss": 3.5496, |
| "step": 20300 |
| }, |
| { |
| "epoch": 2.1902916801205468, |
| "grad_norm": 0.5425307750701904, |
| "learning_rate": 0.00046913263656933515, |
| "loss": 3.5726, |
| "step": 20350 |
| }, |
| { |
| "epoch": 2.195673232160155, |
| "grad_norm": 0.5559353828430176, |
| "learning_rate": 0.00046880939553927374, |
| "loss": 3.5448, |
| "step": 20400 |
| }, |
| { |
| "epoch": 2.2010547841997634, |
| "grad_norm": 0.5595205426216125, |
| "learning_rate": 0.00046848615450921234, |
| "loss": 3.5519, |
| "step": 20450 |
| }, |
| { |
| "epoch": 2.2064363362393715, |
| "grad_norm": 0.5760559439659119, |
| "learning_rate": 0.0004681629134791509, |
| "loss": 3.5624, |
| "step": 20500 |
| }, |
| { |
| "epoch": 2.2118178882789796, |
| "grad_norm": 0.5371535420417786, |
| "learning_rate": 0.00046783967244908947, |
| "loss": 3.5661, |
| "step": 20550 |
| }, |
| { |
| "epoch": 2.2171994403185877, |
| "grad_norm": 0.5587826371192932, |
| "learning_rate": 0.0004675164314190281, |
| "loss": 3.5571, |
| "step": 20600 |
| }, |
| { |
| "epoch": 2.2225809923581963, |
| "grad_norm": 0.6047276258468628, |
| "learning_rate": 0.00046719319038896666, |
| "loss": 3.5486, |
| "step": 20650 |
| }, |
| { |
| "epoch": 2.2279625443978044, |
| "grad_norm": 0.5768632292747498, |
| "learning_rate": 0.00046686994935890526, |
| "loss": 3.5546, |
| "step": 20700 |
| }, |
| { |
| "epoch": 2.2333440964374125, |
| "grad_norm": 0.6263005137443542, |
| "learning_rate": 0.0004665467083288438, |
| "loss": 3.554, |
| "step": 20750 |
| }, |
| { |
| "epoch": 2.2387256484770206, |
| "grad_norm": 0.6004786491394043, |
| "learning_rate": 0.0004662234672987824, |
| "loss": 3.5518, |
| "step": 20800 |
| }, |
| { |
| "epoch": 2.244107200516629, |
| "grad_norm": 0.6552075147628784, |
| "learning_rate": 0.00046590022626872104, |
| "loss": 3.549, |
| "step": 20850 |
| }, |
| { |
| "epoch": 2.2494887525562373, |
| "grad_norm": 0.5540356636047363, |
| "learning_rate": 0.0004655769852386596, |
| "loss": 3.5524, |
| "step": 20900 |
| }, |
| { |
| "epoch": 2.2548703045958454, |
| "grad_norm": 0.593418300151825, |
| "learning_rate": 0.0004652537442085982, |
| "loss": 3.5517, |
| "step": 20950 |
| }, |
| { |
| "epoch": 2.2602518566354535, |
| "grad_norm": 0.5552871823310852, |
| "learning_rate": 0.00046493050317853677, |
| "loss": 3.5791, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.2602518566354535, |
| "eval_accuracy": 0.3637226043966826, |
| "eval_loss": 3.56265926361084, |
| "eval_runtime": 179.7058, |
| "eval_samples_per_second": 100.225, |
| "eval_steps_per_second": 6.266, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.265633408675062, |
| "grad_norm": 0.5663247108459473, |
| "learning_rate": 0.0004646072621484753, |
| "loss": 3.5696, |
| "step": 21050 |
| }, |
| { |
| "epoch": 2.27101496071467, |
| "grad_norm": 0.5453855395317078, |
| "learning_rate": 0.0004642840211184139, |
| "loss": 3.5571, |
| "step": 21100 |
| }, |
| { |
| "epoch": 2.2763965127542782, |
| "grad_norm": 0.5427026748657227, |
| "learning_rate": 0.00046396078008835255, |
| "loss": 3.5705, |
| "step": 21150 |
| }, |
| { |
| "epoch": 2.281778064793887, |
| "grad_norm": 0.9001597166061401, |
| "learning_rate": 0.0004636375390582911, |
| "loss": 3.5635, |
| "step": 21200 |
| }, |
| { |
| "epoch": 2.287159616833495, |
| "grad_norm": 0.5693010687828064, |
| "learning_rate": 0.0004633142980282297, |
| "loss": 3.5494, |
| "step": 21250 |
| }, |
| { |
| "epoch": 2.292541168873103, |
| "grad_norm": 0.5871009230613708, |
| "learning_rate": 0.00046299105699816823, |
| "loss": 3.5662, |
| "step": 21300 |
| }, |
| { |
| "epoch": 2.297922720912711, |
| "grad_norm": 0.5679400563240051, |
| "learning_rate": 0.0004626678159681068, |
| "loss": 3.5698, |
| "step": 21350 |
| }, |
| { |
| "epoch": 2.303304272952319, |
| "grad_norm": 0.5684141516685486, |
| "learning_rate": 0.0004623445749380454, |
| "loss": 3.5673, |
| "step": 21400 |
| }, |
| { |
| "epoch": 2.3086858249919278, |
| "grad_norm": 0.5958677530288696, |
| "learning_rate": 0.000462021333907984, |
| "loss": 3.5621, |
| "step": 21450 |
| }, |
| { |
| "epoch": 2.314067377031536, |
| "grad_norm": 0.5784631967544556, |
| "learning_rate": 0.0004616980928779226, |
| "loss": 3.5556, |
| "step": 21500 |
| }, |
| { |
| "epoch": 2.319448929071144, |
| "grad_norm": 0.5525918006896973, |
| "learning_rate": 0.0004613748518478612, |
| "loss": 3.5443, |
| "step": 21550 |
| }, |
| { |
| "epoch": 2.3248304811107525, |
| "grad_norm": 0.561515212059021, |
| "learning_rate": 0.00046105161081779974, |
| "loss": 3.5642, |
| "step": 21600 |
| }, |
| { |
| "epoch": 2.3302120331503606, |
| "grad_norm": 0.5596020817756653, |
| "learning_rate": 0.00046072836978773834, |
| "loss": 3.5656, |
| "step": 21650 |
| }, |
| { |
| "epoch": 2.3355935851899687, |
| "grad_norm": 0.6286318898200989, |
| "learning_rate": 0.000460405128757677, |
| "loss": 3.5574, |
| "step": 21700 |
| }, |
| { |
| "epoch": 2.340975137229577, |
| "grad_norm": 0.58785080909729, |
| "learning_rate": 0.0004600818877276155, |
| "loss": 3.5552, |
| "step": 21750 |
| }, |
| { |
| "epoch": 2.3463566892691854, |
| "grad_norm": 0.5441408753395081, |
| "learning_rate": 0.0004597586466975541, |
| "loss": 3.5477, |
| "step": 21800 |
| }, |
| { |
| "epoch": 2.3517382413087935, |
| "grad_norm": 0.5354775190353394, |
| "learning_rate": 0.00045943540566749266, |
| "loss": 3.5428, |
| "step": 21850 |
| }, |
| { |
| "epoch": 2.3571197933484016, |
| "grad_norm": 0.5691161155700684, |
| "learning_rate": 0.00045911216463743126, |
| "loss": 3.5492, |
| "step": 21900 |
| }, |
| { |
| "epoch": 2.3625013453880097, |
| "grad_norm": 0.5708732008934021, |
| "learning_rate": 0.00045878892360736985, |
| "loss": 3.568, |
| "step": 21950 |
| }, |
| { |
| "epoch": 2.3678828974276183, |
| "grad_norm": 0.5911089777946472, |
| "learning_rate": 0.00045847214739790966, |
| "loss": 3.5606, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3678828974276183, |
| "eval_accuracy": 0.3649360398477947, |
| "eval_loss": 3.547260284423828, |
| "eval_runtime": 180.1433, |
| "eval_samples_per_second": 99.981, |
| "eval_steps_per_second": 6.251, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3732644494672264, |
| "grad_norm": 0.5441485047340393, |
| "learning_rate": 0.0004581489063678482, |
| "loss": 3.5519, |
| "step": 22050 |
| }, |
| { |
| "epoch": 2.3786460015068345, |
| "grad_norm": 0.561090350151062, |
| "learning_rate": 0.00045782566533778685, |
| "loss": 3.5399, |
| "step": 22100 |
| }, |
| { |
| "epoch": 2.384027553546443, |
| "grad_norm": 0.5627390146255493, |
| "learning_rate": 0.00045750242430772544, |
| "loss": 3.564, |
| "step": 22150 |
| }, |
| { |
| "epoch": 2.389409105586051, |
| "grad_norm": 0.5869255065917969, |
| "learning_rate": 0.000457179183277664, |
| "loss": 3.55, |
| "step": 22200 |
| }, |
| { |
| "epoch": 2.3947906576256592, |
| "grad_norm": 0.6055136322975159, |
| "learning_rate": 0.0004568559422476026, |
| "loss": 3.5617, |
| "step": 22250 |
| }, |
| { |
| "epoch": 2.4001722096652673, |
| "grad_norm": 0.598421037197113, |
| "learning_rate": 0.00045653270121754117, |
| "loss": 3.5477, |
| "step": 22300 |
| }, |
| { |
| "epoch": 2.4055537617048754, |
| "grad_norm": 0.6216960549354553, |
| "learning_rate": 0.0004562094601874797, |
| "loss": 3.5399, |
| "step": 22350 |
| }, |
| { |
| "epoch": 2.410935313744484, |
| "grad_norm": 0.5606902837753296, |
| "learning_rate": 0.00045588621915741836, |
| "loss": 3.5615, |
| "step": 22400 |
| }, |
| { |
| "epoch": 2.416316865784092, |
| "grad_norm": 0.635993242263794, |
| "learning_rate": 0.00045556297812735696, |
| "loss": 3.5505, |
| "step": 22450 |
| }, |
| { |
| "epoch": 2.4216984178237, |
| "grad_norm": 0.5704584121704102, |
| "learning_rate": 0.0004552397370972955, |
| "loss": 3.5438, |
| "step": 22500 |
| }, |
| { |
| "epoch": 2.4270799698633088, |
| "grad_norm": 0.564420759677887, |
| "learning_rate": 0.0004549164960672341, |
| "loss": 3.5614, |
| "step": 22550 |
| }, |
| { |
| "epoch": 2.432461521902917, |
| "grad_norm": 0.5742602348327637, |
| "learning_rate": 0.00045459325503717263, |
| "loss": 3.5337, |
| "step": 22600 |
| }, |
| { |
| "epoch": 2.437843073942525, |
| "grad_norm": 0.5805179476737976, |
| "learning_rate": 0.0004542700140071113, |
| "loss": 3.5488, |
| "step": 22650 |
| }, |
| { |
| "epoch": 2.443224625982133, |
| "grad_norm": 0.5458112359046936, |
| "learning_rate": 0.0004539467729770499, |
| "loss": 3.547, |
| "step": 22700 |
| }, |
| { |
| "epoch": 2.4486061780217416, |
| "grad_norm": 0.5278791189193726, |
| "learning_rate": 0.0004536235319469884, |
| "loss": 3.5493, |
| "step": 22750 |
| }, |
| { |
| "epoch": 2.4539877300613497, |
| "grad_norm": 0.5699906945228577, |
| "learning_rate": 0.000453300290916927, |
| "loss": 3.5378, |
| "step": 22800 |
| }, |
| { |
| "epoch": 2.459369282100958, |
| "grad_norm": 0.5342678427696228, |
| "learning_rate": 0.0004529770498868656, |
| "loss": 3.5485, |
| "step": 22850 |
| }, |
| { |
| "epoch": 2.464750834140566, |
| "grad_norm": 0.5861589908599854, |
| "learning_rate": 0.00045265380885680414, |
| "loss": 3.576, |
| "step": 22900 |
| }, |
| { |
| "epoch": 2.4701323861801745, |
| "grad_norm": 0.5228945016860962, |
| "learning_rate": 0.0004523305678267428, |
| "loss": 3.5571, |
| "step": 22950 |
| }, |
| { |
| "epoch": 2.4755139382197826, |
| "grad_norm": 0.5882893204689026, |
| "learning_rate": 0.0004520073267966814, |
| "loss": 3.5327, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4755139382197826, |
| "eval_accuracy": 0.3658152590159895, |
| "eval_loss": 3.540963888168335, |
| "eval_runtime": 179.7322, |
| "eval_samples_per_second": 100.21, |
| "eval_steps_per_second": 6.265, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4808954902593907, |
| "grad_norm": 0.6533792018890381, |
| "learning_rate": 0.00045168408576661993, |
| "loss": 3.542, |
| "step": 23050 |
| }, |
| { |
| "epoch": 2.4862770422989993, |
| "grad_norm": 0.5604386329650879, |
| "learning_rate": 0.0004513608447365585, |
| "loss": 3.5655, |
| "step": 23100 |
| }, |
| { |
| "epoch": 2.4916585943386074, |
| "grad_norm": 0.6017301678657532, |
| "learning_rate": 0.00045103760370649706, |
| "loss": 3.5436, |
| "step": 23150 |
| }, |
| { |
| "epoch": 2.4970401463782155, |
| "grad_norm": 0.5612083077430725, |
| "learning_rate": 0.00045071436267643566, |
| "loss": 3.5427, |
| "step": 23200 |
| }, |
| { |
| "epoch": 2.5024216984178236, |
| "grad_norm": 0.5622110366821289, |
| "learning_rate": 0.0004503911216463743, |
| "loss": 3.5466, |
| "step": 23250 |
| }, |
| { |
| "epoch": 2.5078032504574317, |
| "grad_norm": 0.5365000367164612, |
| "learning_rate": 0.00045006788061631285, |
| "loss": 3.5615, |
| "step": 23300 |
| }, |
| { |
| "epoch": 2.5131848024970402, |
| "grad_norm": 0.5494401454925537, |
| "learning_rate": 0.00044974463958625144, |
| "loss": 3.5428, |
| "step": 23350 |
| }, |
| { |
| "epoch": 2.5185663545366483, |
| "grad_norm": 0.5880193114280701, |
| "learning_rate": 0.00044942139855619004, |
| "loss": 3.5372, |
| "step": 23400 |
| }, |
| { |
| "epoch": 2.5239479065762565, |
| "grad_norm": 0.520449161529541, |
| "learning_rate": 0.0004490981575261286, |
| "loss": 3.5542, |
| "step": 23450 |
| }, |
| { |
| "epoch": 2.529329458615865, |
| "grad_norm": 0.5952346920967102, |
| "learning_rate": 0.0004487749164960672, |
| "loss": 3.5421, |
| "step": 23500 |
| }, |
| { |
| "epoch": 2.534711010655473, |
| "grad_norm": 0.5716413259506226, |
| "learning_rate": 0.0004484516754660058, |
| "loss": 3.5611, |
| "step": 23550 |
| }, |
| { |
| "epoch": 2.540092562695081, |
| "grad_norm": 0.5482928156852722, |
| "learning_rate": 0.00044812843443594436, |
| "loss": 3.536, |
| "step": 23600 |
| }, |
| { |
| "epoch": 2.5454741147346893, |
| "grad_norm": 0.5782982110977173, |
| "learning_rate": 0.00044780519340588296, |
| "loss": 3.5429, |
| "step": 23650 |
| }, |
| { |
| "epoch": 2.550855666774298, |
| "grad_norm": 0.607822060585022, |
| "learning_rate": 0.0004474819523758215, |
| "loss": 3.5595, |
| "step": 23700 |
| }, |
| { |
| "epoch": 2.556237218813906, |
| "grad_norm": 0.5348266363143921, |
| "learning_rate": 0.0004471587113457601, |
| "loss": 3.5514, |
| "step": 23750 |
| }, |
| { |
| "epoch": 2.561618770853514, |
| "grad_norm": 0.5660938620567322, |
| "learning_rate": 0.00044683547031569874, |
| "loss": 3.5448, |
| "step": 23800 |
| }, |
| { |
| "epoch": 2.567000322893122, |
| "grad_norm": 0.5412974953651428, |
| "learning_rate": 0.0004465122292856373, |
| "loss": 3.5323, |
| "step": 23850 |
| }, |
| { |
| "epoch": 2.5723818749327307, |
| "grad_norm": 0.5600056648254395, |
| "learning_rate": 0.0004461889882555759, |
| "loss": 3.5325, |
| "step": 23900 |
| }, |
| { |
| "epoch": 2.577763426972339, |
| "grad_norm": 0.5820670127868652, |
| "learning_rate": 0.00044586574722551447, |
| "loss": 3.5247, |
| "step": 23950 |
| }, |
| { |
| "epoch": 2.583144979011947, |
| "grad_norm": 0.5861896276473999, |
| "learning_rate": 0.0004455489710160543, |
| "loss": 3.5435, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.583144979011947, |
| "eval_accuracy": 0.3670534673255233, |
| "eval_loss": 3.5298831462860107, |
| "eval_runtime": 180.4928, |
| "eval_samples_per_second": 99.788, |
| "eval_steps_per_second": 6.238, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.5885265310515555, |
| "grad_norm": 0.5437816381454468, |
| "learning_rate": 0.0004452257299859928, |
| "loss": 3.5307, |
| "step": 24050 |
| }, |
| { |
| "epoch": 2.5939080830911636, |
| "grad_norm": 0.5888408422470093, |
| "learning_rate": 0.0004449024889559314, |
| "loss": 3.5398, |
| "step": 24100 |
| }, |
| { |
| "epoch": 2.5992896351307717, |
| "grad_norm": 0.5409483909606934, |
| "learning_rate": 0.00044457924792587, |
| "loss": 3.5417, |
| "step": 24150 |
| }, |
| { |
| "epoch": 2.60467118717038, |
| "grad_norm": 0.5518124103546143, |
| "learning_rate": 0.0004442560068958086, |
| "loss": 3.5487, |
| "step": 24200 |
| }, |
| { |
| "epoch": 2.610052739209988, |
| "grad_norm": 0.5882364511489868, |
| "learning_rate": 0.0004439327658657472, |
| "loss": 3.5487, |
| "step": 24250 |
| }, |
| { |
| "epoch": 2.6154342912495965, |
| "grad_norm": 0.6176393032073975, |
| "learning_rate": 0.0004436095248356858, |
| "loss": 3.5503, |
| "step": 24300 |
| }, |
| { |
| "epoch": 2.6208158432892046, |
| "grad_norm": 0.6049776077270508, |
| "learning_rate": 0.00044328628380562433, |
| "loss": 3.5415, |
| "step": 24350 |
| }, |
| { |
| "epoch": 2.6261973953288127, |
| "grad_norm": 0.585410475730896, |
| "learning_rate": 0.0004429630427755629, |
| "loss": 3.5487, |
| "step": 24400 |
| }, |
| { |
| "epoch": 2.6315789473684212, |
| "grad_norm": 0.5842284560203552, |
| "learning_rate": 0.0004426398017455016, |
| "loss": 3.5549, |
| "step": 24450 |
| }, |
| { |
| "epoch": 2.6369604994080293, |
| "grad_norm": 0.57343590259552, |
| "learning_rate": 0.0004423165607154401, |
| "loss": 3.5565, |
| "step": 24500 |
| }, |
| { |
| "epoch": 2.6423420514476375, |
| "grad_norm": 0.6217878460884094, |
| "learning_rate": 0.0004419933196853787, |
| "loss": 3.5196, |
| "step": 24550 |
| }, |
| { |
| "epoch": 2.6477236034872456, |
| "grad_norm": 0.5150393843650818, |
| "learning_rate": 0.00044167007865531725, |
| "loss": 3.5427, |
| "step": 24600 |
| }, |
| { |
| "epoch": 2.653105155526854, |
| "grad_norm": 0.5708527565002441, |
| "learning_rate": 0.00044134683762525584, |
| "loss": 3.5328, |
| "step": 24650 |
| }, |
| { |
| "epoch": 2.658486707566462, |
| "grad_norm": 0.5596299767494202, |
| "learning_rate": 0.00044102359659519444, |
| "loss": 3.5424, |
| "step": 24700 |
| }, |
| { |
| "epoch": 2.6638682596060703, |
| "grad_norm": 0.5434728860855103, |
| "learning_rate": 0.00044070035556513303, |
| "loss": 3.5557, |
| "step": 24750 |
| }, |
| { |
| "epoch": 2.6692498116456784, |
| "grad_norm": 0.5621779561042786, |
| "learning_rate": 0.00044037711453507163, |
| "loss": 3.5385, |
| "step": 24800 |
| }, |
| { |
| "epoch": 2.674631363685287, |
| "grad_norm": 0.5845193862915039, |
| "learning_rate": 0.0004400538735050102, |
| "loss": 3.5583, |
| "step": 24850 |
| }, |
| { |
| "epoch": 2.680012915724895, |
| "grad_norm": 0.5741780996322632, |
| "learning_rate": 0.00043973063247494876, |
| "loss": 3.5429, |
| "step": 24900 |
| }, |
| { |
| "epoch": 2.685394467764503, |
| "grad_norm": 0.5549833178520203, |
| "learning_rate": 0.00043940739144488736, |
| "loss": 3.5355, |
| "step": 24950 |
| }, |
| { |
| "epoch": 2.6907760198041117, |
| "grad_norm": 0.5826306939125061, |
| "learning_rate": 0.0004390841504148259, |
| "loss": 3.5384, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.6907760198041117, |
| "eval_accuracy": 0.3680882774290684, |
| "eval_loss": 3.5210931301116943, |
| "eval_runtime": 179.7785, |
| "eval_samples_per_second": 100.184, |
| "eval_steps_per_second": 6.263, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.69615757184372, |
| "grad_norm": 0.5777888894081116, |
| "learning_rate": 0.00043876090938476455, |
| "loss": 3.5599, |
| "step": 25050 |
| }, |
| { |
| "epoch": 2.701539123883328, |
| "grad_norm": 0.5324108600616455, |
| "learning_rate": 0.00043843766835470314, |
| "loss": 3.5358, |
| "step": 25100 |
| }, |
| { |
| "epoch": 2.706920675922936, |
| "grad_norm": 0.6265562772750854, |
| "learning_rate": 0.0004381144273246417, |
| "loss": 3.5241, |
| "step": 25150 |
| }, |
| { |
| "epoch": 2.712302227962544, |
| "grad_norm": 0.5996240377426147, |
| "learning_rate": 0.0004377911862945803, |
| "loss": 3.5503, |
| "step": 25200 |
| }, |
| { |
| "epoch": 2.7176837800021527, |
| "grad_norm": 0.5723291635513306, |
| "learning_rate": 0.00043746794526451887, |
| "loss": 3.531, |
| "step": 25250 |
| }, |
| { |
| "epoch": 2.723065332041761, |
| "grad_norm": 0.5591868758201599, |
| "learning_rate": 0.00043714470423445747, |
| "loss": 3.5112, |
| "step": 25300 |
| }, |
| { |
| "epoch": 2.728446884081369, |
| "grad_norm": 0.6413562893867493, |
| "learning_rate": 0.00043682146320439606, |
| "loss": 3.547, |
| "step": 25350 |
| }, |
| { |
| "epoch": 2.7338284361209775, |
| "grad_norm": 0.5785355567932129, |
| "learning_rate": 0.00043649822217433466, |
| "loss": 3.5309, |
| "step": 25400 |
| }, |
| { |
| "epoch": 2.7392099881605856, |
| "grad_norm": 0.5443965792655945, |
| "learning_rate": 0.0004361749811442732, |
| "loss": 3.5263, |
| "step": 25450 |
| }, |
| { |
| "epoch": 2.7445915402001937, |
| "grad_norm": 0.5387240052223206, |
| "learning_rate": 0.0004358517401142118, |
| "loss": 3.528, |
| "step": 25500 |
| }, |
| { |
| "epoch": 2.749973092239802, |
| "grad_norm": 0.5816372632980347, |
| "learning_rate": 0.00043552849908415033, |
| "loss": 3.5582, |
| "step": 25550 |
| }, |
| { |
| "epoch": 2.7553546442794103, |
| "grad_norm": 0.5390015244483948, |
| "learning_rate": 0.000435205258054089, |
| "loss": 3.5369, |
| "step": 25600 |
| }, |
| { |
| "epoch": 2.7607361963190185, |
| "grad_norm": 0.5933911204338074, |
| "learning_rate": 0.0004348820170240276, |
| "loss": 3.5282, |
| "step": 25650 |
| }, |
| { |
| "epoch": 2.7661177483586266, |
| "grad_norm": 0.5716332793235779, |
| "learning_rate": 0.0004345587759939661, |
| "loss": 3.5247, |
| "step": 25700 |
| }, |
| { |
| "epoch": 2.7714993003982347, |
| "grad_norm": 0.5888131856918335, |
| "learning_rate": 0.0004342355349639047, |
| "loss": 3.5367, |
| "step": 25750 |
| }, |
| { |
| "epoch": 2.776880852437843, |
| "grad_norm": 0.5930463671684265, |
| "learning_rate": 0.0004339122939338433, |
| "loss": 3.5348, |
| "step": 25800 |
| }, |
| { |
| "epoch": 2.7822624044774513, |
| "grad_norm": 0.5617492198944092, |
| "learning_rate": 0.00043358905290378184, |
| "loss": 3.5135, |
| "step": 25850 |
| }, |
| { |
| "epoch": 2.7876439565170594, |
| "grad_norm": 0.5550035834312439, |
| "learning_rate": 0.0004332658118737205, |
| "loss": 3.5386, |
| "step": 25900 |
| }, |
| { |
| "epoch": 2.793025508556668, |
| "grad_norm": 0.545727014541626, |
| "learning_rate": 0.0004329425708436591, |
| "loss": 3.5242, |
| "step": 25950 |
| }, |
| { |
| "epoch": 2.798407060596276, |
| "grad_norm": 0.588079035282135, |
| "learning_rate": 0.0004326257946341989, |
| "loss": 3.5324, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.798407060596276, |
| "eval_accuracy": 0.36903094988374685, |
| "eval_loss": 3.512333631515503, |
| "eval_runtime": 180.2641, |
| "eval_samples_per_second": 99.914, |
| "eval_steps_per_second": 6.246, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.803788612635884, |
| "grad_norm": 0.5898637175559998, |
| "learning_rate": 0.00043230255360413744, |
| "loss": 3.5297, |
| "step": 26050 |
| }, |
| { |
| "epoch": 2.8091701646754923, |
| "grad_norm": 0.6078748106956482, |
| "learning_rate": 0.00043197931257407603, |
| "loss": 3.5293, |
| "step": 26100 |
| }, |
| { |
| "epoch": 2.8145517167151004, |
| "grad_norm": 0.5512140989303589, |
| "learning_rate": 0.0004316560715440146, |
| "loss": 3.5403, |
| "step": 26150 |
| }, |
| { |
| "epoch": 2.819933268754709, |
| "grad_norm": 0.5708469748497009, |
| "learning_rate": 0.00043133283051395317, |
| "loss": 3.516, |
| "step": 26200 |
| }, |
| { |
| "epoch": 2.825314820794317, |
| "grad_norm": 0.5742740035057068, |
| "learning_rate": 0.0004310095894838918, |
| "loss": 3.5201, |
| "step": 26250 |
| }, |
| { |
| "epoch": 2.830696372833925, |
| "grad_norm": 0.5645484924316406, |
| "learning_rate": 0.0004306863484538304, |
| "loss": 3.5243, |
| "step": 26300 |
| }, |
| { |
| "epoch": 2.8360779248735337, |
| "grad_norm": 0.5661312341690063, |
| "learning_rate": 0.00043036310742376895, |
| "loss": 3.5189, |
| "step": 26350 |
| }, |
| { |
| "epoch": 2.841459476913142, |
| "grad_norm": 0.5699840784072876, |
| "learning_rate": 0.00043003986639370754, |
| "loss": 3.5436, |
| "step": 26400 |
| }, |
| { |
| "epoch": 2.84684102895275, |
| "grad_norm": 0.5516175031661987, |
| "learning_rate": 0.0004297166253636461, |
| "loss": 3.5463, |
| "step": 26450 |
| }, |
| { |
| "epoch": 2.852222580992358, |
| "grad_norm": 0.5519968271255493, |
| "learning_rate": 0.0004293933843335847, |
| "loss": 3.5243, |
| "step": 26500 |
| }, |
| { |
| "epoch": 2.857604133031966, |
| "grad_norm": 0.5642961859703064, |
| "learning_rate": 0.00042907014330352333, |
| "loss": 3.5221, |
| "step": 26550 |
| }, |
| { |
| "epoch": 2.8629856850715747, |
| "grad_norm": 0.623792290687561, |
| "learning_rate": 0.00042874690227346187, |
| "loss": 3.5376, |
| "step": 26600 |
| }, |
| { |
| "epoch": 2.868367237111183, |
| "grad_norm": 0.5826441049575806, |
| "learning_rate": 0.00042842366124340046, |
| "loss": 3.5282, |
| "step": 26650 |
| }, |
| { |
| "epoch": 2.873748789150791, |
| "grad_norm": 0.5754335522651672, |
| "learning_rate": 0.00042810042021333906, |
| "loss": 3.5277, |
| "step": 26700 |
| }, |
| { |
| "epoch": 2.8791303411903995, |
| "grad_norm": 0.5910266041755676, |
| "learning_rate": 0.0004277771791832776, |
| "loss": 3.5288, |
| "step": 26750 |
| }, |
| { |
| "epoch": 2.8845118932300076, |
| "grad_norm": 0.5650596022605896, |
| "learning_rate": 0.0004274539381532162, |
| "loss": 3.5197, |
| "step": 26800 |
| }, |
| { |
| "epoch": 2.8898934452696157, |
| "grad_norm": 0.6088457703590393, |
| "learning_rate": 0.00042713069712315484, |
| "loss": 3.51, |
| "step": 26850 |
| }, |
| { |
| "epoch": 2.895274997309224, |
| "grad_norm": 0.5383467674255371, |
| "learning_rate": 0.0004268074560930934, |
| "loss": 3.5174, |
| "step": 26900 |
| }, |
| { |
| "epoch": 2.9006565493488323, |
| "grad_norm": 0.5610182881355286, |
| "learning_rate": 0.000426484215063032, |
| "loss": 3.5204, |
| "step": 26950 |
| }, |
| { |
| "epoch": 2.9060381013884404, |
| "grad_norm": 0.5684370994567871, |
| "learning_rate": 0.0004261609740329705, |
| "loss": 3.5153, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9060381013884404, |
| "eval_accuracy": 0.36983009187362237, |
| "eval_loss": 3.501441717147827, |
| "eval_runtime": 179.9533, |
| "eval_samples_per_second": 100.087, |
| "eval_steps_per_second": 6.257, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9114196534280485, |
| "grad_norm": 0.6579453349113464, |
| "learning_rate": 0.0004258377330029091, |
| "loss": 3.5223, |
| "step": 27050 |
| }, |
| { |
| "epoch": 2.9168012054676566, |
| "grad_norm": 0.5974162220954895, |
| "learning_rate": 0.00042551449197284776, |
| "loss": 3.5281, |
| "step": 27100 |
| }, |
| { |
| "epoch": 2.922182757507265, |
| "grad_norm": 0.5756824016571045, |
| "learning_rate": 0.0004251912509427863, |
| "loss": 3.5124, |
| "step": 27150 |
| }, |
| { |
| "epoch": 2.9275643095468733, |
| "grad_norm": 0.639370858669281, |
| "learning_rate": 0.0004248680099127249, |
| "loss": 3.5413, |
| "step": 27200 |
| }, |
| { |
| "epoch": 2.9329458615864814, |
| "grad_norm": 0.597856879234314, |
| "learning_rate": 0.0004245447688826635, |
| "loss": 3.5282, |
| "step": 27250 |
| }, |
| { |
| "epoch": 2.93832741362609, |
| "grad_norm": 0.6288902163505554, |
| "learning_rate": 0.00042422152785260203, |
| "loss": 3.532, |
| "step": 27300 |
| }, |
| { |
| "epoch": 2.943708965665698, |
| "grad_norm": 0.7320722937583923, |
| "learning_rate": 0.0004238982868225406, |
| "loss": 3.5306, |
| "step": 27350 |
| }, |
| { |
| "epoch": 2.949090517705306, |
| "grad_norm": 0.5934067964553833, |
| "learning_rate": 0.0004235750457924793, |
| "loss": 3.5274, |
| "step": 27400 |
| }, |
| { |
| "epoch": 2.9544720697449143, |
| "grad_norm": 0.554151177406311, |
| "learning_rate": 0.0004232518047624178, |
| "loss": 3.5108, |
| "step": 27450 |
| }, |
| { |
| "epoch": 2.9598536217845224, |
| "grad_norm": 0.552830159664154, |
| "learning_rate": 0.0004229285637323564, |
| "loss": 3.515, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.965235173824131, |
| "grad_norm": 0.5718715190887451, |
| "learning_rate": 0.00042260532270229495, |
| "loss": 3.5394, |
| "step": 27550 |
| }, |
| { |
| "epoch": 2.970616725863739, |
| "grad_norm": 0.5800749063491821, |
| "learning_rate": 0.00042228208167223354, |
| "loss": 3.5282, |
| "step": 27600 |
| }, |
| { |
| "epoch": 2.975998277903347, |
| "grad_norm": 0.6078159809112549, |
| "learning_rate": 0.00042195884064217214, |
| "loss": 3.5321, |
| "step": 27650 |
| }, |
| { |
| "epoch": 2.9813798299429557, |
| "grad_norm": 0.5575968623161316, |
| "learning_rate": 0.00042163559961211073, |
| "loss": 3.524, |
| "step": 27700 |
| }, |
| { |
| "epoch": 2.986761381982564, |
| "grad_norm": 0.5276609659194946, |
| "learning_rate": 0.00042131235858204933, |
| "loss": 3.5273, |
| "step": 27750 |
| }, |
| { |
| "epoch": 2.992142934022172, |
| "grad_norm": 0.5870263576507568, |
| "learning_rate": 0.0004209891175519879, |
| "loss": 3.5096, |
| "step": 27800 |
| }, |
| { |
| "epoch": 2.9975244860617805, |
| "grad_norm": 0.5546600818634033, |
| "learning_rate": 0.00042066587652192646, |
| "loss": 3.5217, |
| "step": 27850 |
| }, |
| { |
| "epoch": 3.0029060381013886, |
| "grad_norm": 0.5607388019561768, |
| "learning_rate": 0.00042034263549186506, |
| "loss": 3.46, |
| "step": 27900 |
| }, |
| { |
| "epoch": 3.0082875901409967, |
| "grad_norm": 0.6114853620529175, |
| "learning_rate": 0.0004200193944618036, |
| "loss": 3.4226, |
| "step": 27950 |
| }, |
| { |
| "epoch": 3.0136691421806048, |
| "grad_norm": 0.5920767188072205, |
| "learning_rate": 0.00041970261825234346, |
| "loss": 3.4362, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.0136691421806048, |
| "eval_accuracy": 0.3704642987797955, |
| "eval_loss": 3.497429132461548, |
| "eval_runtime": 180.2708, |
| "eval_samples_per_second": 99.911, |
| "eval_steps_per_second": 6.246, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.0190506942202133, |
| "grad_norm": 0.6086764335632324, |
| "learning_rate": 0.00041937937722228205, |
| "loss": 3.4374, |
| "step": 28050 |
| }, |
| { |
| "epoch": 3.0244322462598214, |
| "grad_norm": 0.573062539100647, |
| "learning_rate": 0.00041905613619222065, |
| "loss": 3.4365, |
| "step": 28100 |
| }, |
| { |
| "epoch": 3.0298137982994295, |
| "grad_norm": 0.6111119389533997, |
| "learning_rate": 0.00041873289516215924, |
| "loss": 3.4324, |
| "step": 28150 |
| }, |
| { |
| "epoch": 3.0351953503390376, |
| "grad_norm": 0.5908578634262085, |
| "learning_rate": 0.0004184096541320978, |
| "loss": 3.4512, |
| "step": 28200 |
| }, |
| { |
| "epoch": 3.040576902378646, |
| "grad_norm": 0.8571264147758484, |
| "learning_rate": 0.0004180864131020364, |
| "loss": 3.4453, |
| "step": 28250 |
| }, |
| { |
| "epoch": 3.0459584544182543, |
| "grad_norm": 0.5717503428459167, |
| "learning_rate": 0.0004177631720719749, |
| "loss": 3.4228, |
| "step": 28300 |
| }, |
| { |
| "epoch": 3.0513400064578624, |
| "grad_norm": 0.606881320476532, |
| "learning_rate": 0.00041743993104191357, |
| "loss": 3.4489, |
| "step": 28350 |
| }, |
| { |
| "epoch": 3.0567215584974705, |
| "grad_norm": 0.579672634601593, |
| "learning_rate": 0.00041711669001185216, |
| "loss": 3.4317, |
| "step": 28400 |
| }, |
| { |
| "epoch": 3.062103110537079, |
| "grad_norm": 0.6016443371772766, |
| "learning_rate": 0.0004167934489817907, |
| "loss": 3.4341, |
| "step": 28450 |
| }, |
| { |
| "epoch": 3.067484662576687, |
| "grad_norm": 0.5978291034698486, |
| "learning_rate": 0.0004164702079517293, |
| "loss": 3.4344, |
| "step": 28500 |
| }, |
| { |
| "epoch": 3.0728662146162953, |
| "grad_norm": 0.5813153386116028, |
| "learning_rate": 0.0004161469669216679, |
| "loss": 3.4579, |
| "step": 28550 |
| }, |
| { |
| "epoch": 3.0782477666559034, |
| "grad_norm": 0.5806988477706909, |
| "learning_rate": 0.00041582372589160643, |
| "loss": 3.4485, |
| "step": 28600 |
| }, |
| { |
| "epoch": 3.083629318695512, |
| "grad_norm": 0.5982168316841125, |
| "learning_rate": 0.0004155004848615451, |
| "loss": 3.4556, |
| "step": 28650 |
| }, |
| { |
| "epoch": 3.08901087073512, |
| "grad_norm": 0.6008811593055725, |
| "learning_rate": 0.0004151772438314837, |
| "loss": 3.4323, |
| "step": 28700 |
| }, |
| { |
| "epoch": 3.094392422774728, |
| "grad_norm": 0.5557597875595093, |
| "learning_rate": 0.0004148540028014222, |
| "loss": 3.4236, |
| "step": 28750 |
| }, |
| { |
| "epoch": 3.0997739748143363, |
| "grad_norm": 0.584446370601654, |
| "learning_rate": 0.0004145307617713608, |
| "loss": 3.4504, |
| "step": 28800 |
| }, |
| { |
| "epoch": 3.105155526853945, |
| "grad_norm": 0.576360821723938, |
| "learning_rate": 0.00041420752074129935, |
| "loss": 3.4543, |
| "step": 28850 |
| }, |
| { |
| "epoch": 3.110537078893553, |
| "grad_norm": 0.5923529267311096, |
| "learning_rate": 0.000413884279711238, |
| "loss": 3.4149, |
| "step": 28900 |
| }, |
| { |
| "epoch": 3.115918630933161, |
| "grad_norm": 0.5620018243789673, |
| "learning_rate": 0.00041356750350177775, |
| "loss": 3.4458, |
| "step": 28950 |
| }, |
| { |
| "epoch": 3.121300182972769, |
| "grad_norm": 0.658657968044281, |
| "learning_rate": 0.0004132442624717164, |
| "loss": 3.448, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.121300182972769, |
| "eval_accuracy": 0.371386979103116, |
| "eval_loss": 3.4938414096832275, |
| "eval_runtime": 180.2028, |
| "eval_samples_per_second": 99.948, |
| "eval_steps_per_second": 6.249, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1266817350123777, |
| "grad_norm": 0.5916072726249695, |
| "learning_rate": 0.000412921021441655, |
| "loss": 3.4481, |
| "step": 29050 |
| }, |
| { |
| "epoch": 3.132063287051986, |
| "grad_norm": 0.5693389177322388, |
| "learning_rate": 0.00041259778041159354, |
| "loss": 3.4506, |
| "step": 29100 |
| }, |
| { |
| "epoch": 3.137444839091594, |
| "grad_norm": 0.5854905247688293, |
| "learning_rate": 0.00041227453938153213, |
| "loss": 3.4614, |
| "step": 29150 |
| }, |
| { |
| "epoch": 3.1428263911312024, |
| "grad_norm": 0.5985913276672363, |
| "learning_rate": 0.00041195129835147067, |
| "loss": 3.4455, |
| "step": 29200 |
| }, |
| { |
| "epoch": 3.1482079431708105, |
| "grad_norm": 0.5984156727790833, |
| "learning_rate": 0.00041162805732140927, |
| "loss": 3.447, |
| "step": 29250 |
| }, |
| { |
| "epoch": 3.1535894952104186, |
| "grad_norm": 0.5744922757148743, |
| "learning_rate": 0.0004113048162913479, |
| "loss": 3.4795, |
| "step": 29300 |
| }, |
| { |
| "epoch": 3.1589710472500268, |
| "grad_norm": 0.6045938730239868, |
| "learning_rate": 0.00041098157526128646, |
| "loss": 3.4548, |
| "step": 29350 |
| }, |
| { |
| "epoch": 3.1643525992896353, |
| "grad_norm": 0.6241233348846436, |
| "learning_rate": 0.00041065833423122505, |
| "loss": 3.4531, |
| "step": 29400 |
| }, |
| { |
| "epoch": 3.1697341513292434, |
| "grad_norm": 0.5867751836776733, |
| "learning_rate": 0.00041033509320116365, |
| "loss": 3.441, |
| "step": 29450 |
| }, |
| { |
| "epoch": 3.1751157033688515, |
| "grad_norm": 0.6040583252906799, |
| "learning_rate": 0.0004100118521711022, |
| "loss": 3.461, |
| "step": 29500 |
| }, |
| { |
| "epoch": 3.1804972554084596, |
| "grad_norm": 0.6353616118431091, |
| "learning_rate": 0.0004096886111410408, |
| "loss": 3.4538, |
| "step": 29550 |
| }, |
| { |
| "epoch": 3.185878807448068, |
| "grad_norm": 0.609554648399353, |
| "learning_rate": 0.00040936537011097943, |
| "loss": 3.4705, |
| "step": 29600 |
| }, |
| { |
| "epoch": 3.1912603594876763, |
| "grad_norm": 0.5806785821914673, |
| "learning_rate": 0.00040904212908091797, |
| "loss": 3.444, |
| "step": 29650 |
| }, |
| { |
| "epoch": 3.1966419115272844, |
| "grad_norm": 0.7444307804107666, |
| "learning_rate": 0.00040871888805085656, |
| "loss": 3.4575, |
| "step": 29700 |
| }, |
| { |
| "epoch": 3.2020234635668925, |
| "grad_norm": 0.6058987379074097, |
| "learning_rate": 0.0004083956470207951, |
| "loss": 3.4707, |
| "step": 29750 |
| }, |
| { |
| "epoch": 3.207405015606501, |
| "grad_norm": 0.5647191405296326, |
| "learning_rate": 0.0004080724059907337, |
| "loss": 3.4502, |
| "step": 29800 |
| }, |
| { |
| "epoch": 3.212786567646109, |
| "grad_norm": 0.6401323080062866, |
| "learning_rate": 0.00040774916496067235, |
| "loss": 3.4423, |
| "step": 29850 |
| }, |
| { |
| "epoch": 3.2181681196857173, |
| "grad_norm": 0.6080923676490784, |
| "learning_rate": 0.0004074259239306109, |
| "loss": 3.4446, |
| "step": 29900 |
| }, |
| { |
| "epoch": 3.2235496717253254, |
| "grad_norm": 0.5872158408164978, |
| "learning_rate": 0.0004071026829005495, |
| "loss": 3.4658, |
| "step": 29950 |
| }, |
| { |
| "epoch": 3.228931223764934, |
| "grad_norm": 0.5680561065673828, |
| "learning_rate": 0.0004067794418704881, |
| "loss": 3.4482, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.228931223764934, |
| "eval_accuracy": 0.3716701285287606, |
| "eval_loss": 3.4923532009124756, |
| "eval_runtime": 180.3398, |
| "eval_samples_per_second": 99.873, |
| "eval_steps_per_second": 6.244, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.234312775804542, |
| "grad_norm": 0.6124100685119629, |
| "learning_rate": 0.0004064562008404266, |
| "loss": 3.4498, |
| "step": 30050 |
| }, |
| { |
| "epoch": 3.23969432784415, |
| "grad_norm": 0.6356387138366699, |
| "learning_rate": 0.0004061329598103652, |
| "loss": 3.4562, |
| "step": 30100 |
| }, |
| { |
| "epoch": 3.2450758798837587, |
| "grad_norm": 0.6179471611976624, |
| "learning_rate": 0.00040580971878030386, |
| "loss": 3.4407, |
| "step": 30150 |
| }, |
| { |
| "epoch": 3.250457431923367, |
| "grad_norm": 0.5960007905960083, |
| "learning_rate": 0.0004054864777502424, |
| "loss": 3.4524, |
| "step": 30200 |
| }, |
| { |
| "epoch": 3.255838983962975, |
| "grad_norm": 0.6349455118179321, |
| "learning_rate": 0.000405163236720181, |
| "loss": 3.4605, |
| "step": 30250 |
| }, |
| { |
| "epoch": 3.261220536002583, |
| "grad_norm": 0.5955315232276917, |
| "learning_rate": 0.00040483999569011954, |
| "loss": 3.4601, |
| "step": 30300 |
| }, |
| { |
| "epoch": 3.2666020880421915, |
| "grad_norm": 0.6215502023696899, |
| "learning_rate": 0.00040451675466005813, |
| "loss": 3.4626, |
| "step": 30350 |
| }, |
| { |
| "epoch": 3.2719836400817996, |
| "grad_norm": 0.5687872171401978, |
| "learning_rate": 0.0004041935136299967, |
| "loss": 3.4495, |
| "step": 30400 |
| }, |
| { |
| "epoch": 3.2773651921214078, |
| "grad_norm": 0.6149540543556213, |
| "learning_rate": 0.0004038702725999353, |
| "loss": 3.4654, |
| "step": 30450 |
| }, |
| { |
| "epoch": 3.282746744161016, |
| "grad_norm": 0.5696414709091187, |
| "learning_rate": 0.0004035470315698739, |
| "loss": 3.4482, |
| "step": 30500 |
| }, |
| { |
| "epoch": 3.2881282962006244, |
| "grad_norm": 0.5921586751937866, |
| "learning_rate": 0.0004032237905398125, |
| "loss": 3.4423, |
| "step": 30550 |
| }, |
| { |
| "epoch": 3.2935098482402325, |
| "grad_norm": 0.6276850700378418, |
| "learning_rate": 0.00040290054950975105, |
| "loss": 3.4592, |
| "step": 30600 |
| }, |
| { |
| "epoch": 3.2988914002798406, |
| "grad_norm": 0.597515344619751, |
| "learning_rate": 0.00040257730847968965, |
| "loss": 3.4615, |
| "step": 30650 |
| }, |
| { |
| "epoch": 3.304272952319449, |
| "grad_norm": 0.6534189581871033, |
| "learning_rate": 0.0004022540674496283, |
| "loss": 3.465, |
| "step": 30700 |
| }, |
| { |
| "epoch": 3.3096545043590573, |
| "grad_norm": 0.6232941746711731, |
| "learning_rate": 0.00040193082641956684, |
| "loss": 3.4645, |
| "step": 30750 |
| }, |
| { |
| "epoch": 3.3150360563986654, |
| "grad_norm": 0.6439439654350281, |
| "learning_rate": 0.00040160758538950543, |
| "loss": 3.4445, |
| "step": 30800 |
| }, |
| { |
| "epoch": 3.3204176084382735, |
| "grad_norm": 0.6507943868637085, |
| "learning_rate": 0.00040128434435944397, |
| "loss": 3.4739, |
| "step": 30850 |
| }, |
| { |
| "epoch": 3.3257991604778816, |
| "grad_norm": 0.6493215560913086, |
| "learning_rate": 0.00040096110332938257, |
| "loss": 3.4431, |
| "step": 30900 |
| }, |
| { |
| "epoch": 3.33118071251749, |
| "grad_norm": 0.6146352291107178, |
| "learning_rate": 0.00040063786229932116, |
| "loss": 3.4583, |
| "step": 30950 |
| }, |
| { |
| "epoch": 3.3365622645570983, |
| "grad_norm": 0.6426176428794861, |
| "learning_rate": 0.00040031462126925975, |
| "loss": 3.4714, |
| "step": 31000 |
| }, |
| { |
| "epoch": 3.3365622645570983, |
| "eval_accuracy": 0.3725597783741855, |
| "eval_loss": 3.482445240020752, |
| "eval_runtime": 179.9242, |
| "eval_samples_per_second": 100.103, |
| "eval_steps_per_second": 6.258, |
| "step": 31000 |
| }, |
| { |
| "epoch": 3.3419438165967064, |
| "grad_norm": 0.6456658244132996, |
| "learning_rate": 0.00039999138023919835, |
| "loss": 3.4466, |
| "step": 31050 |
| }, |
| { |
| "epoch": 3.347325368636315, |
| "grad_norm": 0.6010160446166992, |
| "learning_rate": 0.00039966813920913694, |
| "loss": 3.4282, |
| "step": 31100 |
| }, |
| { |
| "epoch": 3.352706920675923, |
| "grad_norm": 0.6064096689224243, |
| "learning_rate": 0.0003993448981790755, |
| "loss": 3.4615, |
| "step": 31150 |
| }, |
| { |
| "epoch": 3.358088472715531, |
| "grad_norm": 0.6383664011955261, |
| "learning_rate": 0.0003990216571490141, |
| "loss": 3.454, |
| "step": 31200 |
| }, |
| { |
| "epoch": 3.3634700247551392, |
| "grad_norm": 0.6114600896835327, |
| "learning_rate": 0.0003986984161189526, |
| "loss": 3.4583, |
| "step": 31250 |
| }, |
| { |
| "epoch": 3.368851576794748, |
| "grad_norm": 0.6199226975440979, |
| "learning_rate": 0.00039837517508889127, |
| "loss": 3.447, |
| "step": 31300 |
| }, |
| { |
| "epoch": 3.374233128834356, |
| "grad_norm": 0.6489754915237427, |
| "learning_rate": 0.00039805193405882986, |
| "loss": 3.4586, |
| "step": 31350 |
| }, |
| { |
| "epoch": 3.379614680873964, |
| "grad_norm": 0.57906174659729, |
| "learning_rate": 0.0003977286930287684, |
| "loss": 3.459, |
| "step": 31400 |
| }, |
| { |
| "epoch": 3.384996232913572, |
| "grad_norm": 0.5564760565757751, |
| "learning_rate": 0.000397405451998707, |
| "loss": 3.4716, |
| "step": 31450 |
| }, |
| { |
| "epoch": 3.3903777849531807, |
| "grad_norm": 0.6209973096847534, |
| "learning_rate": 0.00039708221096864554, |
| "loss": 3.471, |
| "step": 31500 |
| }, |
| { |
| "epoch": 3.3957593369927888, |
| "grad_norm": 0.5805020928382874, |
| "learning_rate": 0.00039675896993858413, |
| "loss": 3.4692, |
| "step": 31550 |
| }, |
| { |
| "epoch": 3.401140889032397, |
| "grad_norm": 0.598192572593689, |
| "learning_rate": 0.0003964357289085228, |
| "loss": 3.4498, |
| "step": 31600 |
| }, |
| { |
| "epoch": 3.4065224410720054, |
| "grad_norm": 0.6662253737449646, |
| "learning_rate": 0.0003961124878784613, |
| "loss": 3.4412, |
| "step": 31650 |
| }, |
| { |
| "epoch": 3.4119039931116135, |
| "grad_norm": 0.5580045580863953, |
| "learning_rate": 0.0003957892468483999, |
| "loss": 3.4489, |
| "step": 31700 |
| }, |
| { |
| "epoch": 3.4172855451512216, |
| "grad_norm": 0.6119815111160278, |
| "learning_rate": 0.0003954660058183385, |
| "loss": 3.4623, |
| "step": 31750 |
| }, |
| { |
| "epoch": 3.4226670971908297, |
| "grad_norm": 0.5958937406539917, |
| "learning_rate": 0.00039514276478827705, |
| "loss": 3.4371, |
| "step": 31800 |
| }, |
| { |
| "epoch": 3.428048649230438, |
| "grad_norm": 0.6033514738082886, |
| "learning_rate": 0.0003948195237582157, |
| "loss": 3.456, |
| "step": 31850 |
| }, |
| { |
| "epoch": 3.4334302012700464, |
| "grad_norm": 0.5335016846656799, |
| "learning_rate": 0.0003944962827281543, |
| "loss": 3.4487, |
| "step": 31900 |
| }, |
| { |
| "epoch": 3.4388117533096545, |
| "grad_norm": 0.6240798234939575, |
| "learning_rate": 0.00039417304169809284, |
| "loss": 3.4618, |
| "step": 31950 |
| }, |
| { |
| "epoch": 3.4441933053492626, |
| "grad_norm": 0.6087186932563782, |
| "learning_rate": 0.00039384980066803143, |
| "loss": 3.4663, |
| "step": 32000 |
| }, |
| { |
| "epoch": 3.4441933053492626, |
| "eval_accuracy": 0.3730128609163719, |
| "eval_loss": 3.4758286476135254, |
| "eval_runtime": 180.9046, |
| "eval_samples_per_second": 99.561, |
| "eval_steps_per_second": 6.224, |
| "step": 32000 |
| }, |
| { |
| "epoch": 3.449574857388871, |
| "grad_norm": 0.5830413103103638, |
| "learning_rate": 0.00039352655963796997, |
| "loss": 3.4582, |
| "step": 32050 |
| }, |
| { |
| "epoch": 3.4549564094284793, |
| "grad_norm": 0.6102946996688843, |
| "learning_rate": 0.00039320331860790857, |
| "loss": 3.4796, |
| "step": 32100 |
| }, |
| { |
| "epoch": 3.4603379614680874, |
| "grad_norm": 0.5744568705558777, |
| "learning_rate": 0.0003928800775778472, |
| "loss": 3.448, |
| "step": 32150 |
| }, |
| { |
| "epoch": 3.4657195135076955, |
| "grad_norm": 0.6026392579078674, |
| "learning_rate": 0.00039255683654778576, |
| "loss": 3.451, |
| "step": 32200 |
| }, |
| { |
| "epoch": 3.471101065547304, |
| "grad_norm": 0.6402329206466675, |
| "learning_rate": 0.00039223359551772435, |
| "loss": 3.4786, |
| "step": 32250 |
| }, |
| { |
| "epoch": 3.476482617586912, |
| "grad_norm": 0.6022635102272034, |
| "learning_rate": 0.00039191035448766294, |
| "loss": 3.461, |
| "step": 32300 |
| }, |
| { |
| "epoch": 3.4818641696265202, |
| "grad_norm": 0.6575260758399963, |
| "learning_rate": 0.0003915871134576015, |
| "loss": 3.4667, |
| "step": 32350 |
| }, |
| { |
| "epoch": 3.4872457216661283, |
| "grad_norm": 0.6149677634239197, |
| "learning_rate": 0.0003912638724275401, |
| "loss": 3.4635, |
| "step": 32400 |
| }, |
| { |
| "epoch": 3.492627273705737, |
| "grad_norm": 0.7180488705635071, |
| "learning_rate": 0.00039094063139747873, |
| "loss": 3.4736, |
| "step": 32450 |
| }, |
| { |
| "epoch": 3.498008825745345, |
| "grad_norm": 0.6193327903747559, |
| "learning_rate": 0.00039061739036741727, |
| "loss": 3.4594, |
| "step": 32500 |
| }, |
| { |
| "epoch": 3.503390377784953, |
| "grad_norm": 0.6941162943840027, |
| "learning_rate": 0.00039029414933735586, |
| "loss": 3.4656, |
| "step": 32550 |
| }, |
| { |
| "epoch": 3.5087719298245617, |
| "grad_norm": 0.5805447101593018, |
| "learning_rate": 0.0003899709083072944, |
| "loss": 3.4647, |
| "step": 32600 |
| }, |
| { |
| "epoch": 3.5141534818641698, |
| "grad_norm": 0.6239206790924072, |
| "learning_rate": 0.000389647667277233, |
| "loss": 3.4568, |
| "step": 32650 |
| }, |
| { |
| "epoch": 3.519535033903778, |
| "grad_norm": 0.6483040452003479, |
| "learning_rate": 0.00038932442624717165, |
| "loss": 3.4532, |
| "step": 32700 |
| }, |
| { |
| "epoch": 3.524916585943386, |
| "grad_norm": 0.6193462014198303, |
| "learning_rate": 0.0003890011852171102, |
| "loss": 3.4473, |
| "step": 32750 |
| }, |
| { |
| "epoch": 3.530298137982994, |
| "grad_norm": 0.6228381991386414, |
| "learning_rate": 0.0003886779441870488, |
| "loss": 3.4511, |
| "step": 32800 |
| }, |
| { |
| "epoch": 3.5356796900226026, |
| "grad_norm": 0.5783094167709351, |
| "learning_rate": 0.0003883547031569874, |
| "loss": 3.444, |
| "step": 32850 |
| }, |
| { |
| "epoch": 3.5410612420622107, |
| "grad_norm": 0.5559167861938477, |
| "learning_rate": 0.0003880314621269259, |
| "loss": 3.4532, |
| "step": 32900 |
| }, |
| { |
| "epoch": 3.546442794101819, |
| "grad_norm": 0.603102445602417, |
| "learning_rate": 0.0003877146859174657, |
| "loss": 3.4626, |
| "step": 32950 |
| }, |
| { |
| "epoch": 3.5518243461414274, |
| "grad_norm": 0.5790406465530396, |
| "learning_rate": 0.0003873914448874043, |
| "loss": 3.461, |
| "step": 33000 |
| }, |
| { |
| "epoch": 3.5518243461414274, |
| "eval_accuracy": 0.3738632870692959, |
| "eval_loss": 3.4701783657073975, |
| "eval_runtime": 180.005, |
| "eval_samples_per_second": 100.058, |
| "eval_steps_per_second": 6.255, |
| "step": 33000 |
| }, |
| { |
| "epoch": 3.5572058981810355, |
| "grad_norm": 0.6018248796463013, |
| "learning_rate": 0.0003870682038573429, |
| "loss": 3.4504, |
| "step": 33050 |
| }, |
| { |
| "epoch": 3.5625874502206436, |
| "grad_norm": 0.5604973435401917, |
| "learning_rate": 0.0003867449628272815, |
| "loss": 3.4534, |
| "step": 33100 |
| }, |
| { |
| "epoch": 3.5679690022602517, |
| "grad_norm": 0.5984012484550476, |
| "learning_rate": 0.0003864217217972201, |
| "loss": 3.4518, |
| "step": 33150 |
| }, |
| { |
| "epoch": 3.57335055429986, |
| "grad_norm": 0.6406269669532776, |
| "learning_rate": 0.0003860984807671587, |
| "loss": 3.4586, |
| "step": 33200 |
| }, |
| { |
| "epoch": 3.5787321063394684, |
| "grad_norm": 0.6193456053733826, |
| "learning_rate": 0.00038577523973709724, |
| "loss": 3.4617, |
| "step": 33250 |
| }, |
| { |
| "epoch": 3.5841136583790765, |
| "grad_norm": 0.6093300580978394, |
| "learning_rate": 0.00038545199870703583, |
| "loss": 3.4654, |
| "step": 33300 |
| }, |
| { |
| "epoch": 3.5894952104186846, |
| "grad_norm": 0.5693668723106384, |
| "learning_rate": 0.0003851287576769744, |
| "loss": 3.4363, |
| "step": 33350 |
| }, |
| { |
| "epoch": 3.594876762458293, |
| "grad_norm": 0.6020606160163879, |
| "learning_rate": 0.000384805516646913, |
| "loss": 3.4486, |
| "step": 33400 |
| }, |
| { |
| "epoch": 3.6002583144979012, |
| "grad_norm": 0.5911715030670166, |
| "learning_rate": 0.0003844822756168516, |
| "loss": 3.4621, |
| "step": 33450 |
| }, |
| { |
| "epoch": 3.6056398665375093, |
| "grad_norm": 0.6165992021560669, |
| "learning_rate": 0.00038415903458679016, |
| "loss": 3.456, |
| "step": 33500 |
| }, |
| { |
| "epoch": 3.611021418577118, |
| "grad_norm": 0.6295124292373657, |
| "learning_rate": 0.00038383579355672875, |
| "loss": 3.4562, |
| "step": 33550 |
| }, |
| { |
| "epoch": 3.616402970616726, |
| "grad_norm": 0.5734518766403198, |
| "learning_rate": 0.00038351255252666735, |
| "loss": 3.4466, |
| "step": 33600 |
| }, |
| { |
| "epoch": 3.621784522656334, |
| "grad_norm": 0.6045584082603455, |
| "learning_rate": 0.00038318931149660594, |
| "loss": 3.4463, |
| "step": 33650 |
| }, |
| { |
| "epoch": 3.627166074695942, |
| "grad_norm": 0.625198483467102, |
| "learning_rate": 0.00038286607046654454, |
| "loss": 3.4344, |
| "step": 33700 |
| }, |
| { |
| "epoch": 3.6325476267355503, |
| "grad_norm": 0.6212444305419922, |
| "learning_rate": 0.00038254282943648313, |
| "loss": 3.4403, |
| "step": 33750 |
| }, |
| { |
| "epoch": 3.637929178775159, |
| "grad_norm": 0.6339898109436035, |
| "learning_rate": 0.00038221958840642167, |
| "loss": 3.4669, |
| "step": 33800 |
| }, |
| { |
| "epoch": 3.643310730814767, |
| "grad_norm": 0.5889163017272949, |
| "learning_rate": 0.00038189634737636027, |
| "loss": 3.4513, |
| "step": 33850 |
| }, |
| { |
| "epoch": 3.648692282854375, |
| "grad_norm": 0.5536656975746155, |
| "learning_rate": 0.0003815731063462988, |
| "loss": 3.4398, |
| "step": 33900 |
| }, |
| { |
| "epoch": 3.6540738348939836, |
| "grad_norm": 0.5549130439758301, |
| "learning_rate": 0.00038124986531623745, |
| "loss": 3.4713, |
| "step": 33950 |
| }, |
| { |
| "epoch": 3.6594553869335917, |
| "grad_norm": 0.6131840944290161, |
| "learning_rate": 0.00038092662428617605, |
| "loss": 3.4507, |
| "step": 34000 |
| }, |
| { |
| "epoch": 3.6594553869335917, |
| "eval_accuracy": 0.37475293691472084, |
| "eval_loss": 3.4609808921813965, |
| "eval_runtime": 180.1311, |
| "eval_samples_per_second": 99.988, |
| "eval_steps_per_second": 6.251, |
| "step": 34000 |
| }, |
| { |
| "epoch": 3.6648369389732, |
| "grad_norm": 0.5674228072166443, |
| "learning_rate": 0.0003806033832561146, |
| "loss": 3.4384, |
| "step": 34050 |
| }, |
| { |
| "epoch": 3.670218491012808, |
| "grad_norm": 0.6528271436691284, |
| "learning_rate": 0.0003802801422260532, |
| "loss": 3.4544, |
| "step": 34100 |
| }, |
| { |
| "epoch": 3.675600043052416, |
| "grad_norm": 0.59865802526474, |
| "learning_rate": 0.0003799569011959918, |
| "loss": 3.4432, |
| "step": 34150 |
| }, |
| { |
| "epoch": 3.6809815950920246, |
| "grad_norm": 0.6160247325897217, |
| "learning_rate": 0.0003796336601659303, |
| "loss": 3.4473, |
| "step": 34200 |
| }, |
| { |
| "epoch": 3.6863631471316327, |
| "grad_norm": 0.6500553488731384, |
| "learning_rate": 0.00037931041913586897, |
| "loss": 3.4589, |
| "step": 34250 |
| }, |
| { |
| "epoch": 3.691744699171241, |
| "grad_norm": 0.6018146276473999, |
| "learning_rate": 0.00037898717810580756, |
| "loss": 3.4633, |
| "step": 34300 |
| }, |
| { |
| "epoch": 3.6971262512108494, |
| "grad_norm": 0.614810585975647, |
| "learning_rate": 0.0003786639370757461, |
| "loss": 3.4687, |
| "step": 34350 |
| }, |
| { |
| "epoch": 3.7025078032504575, |
| "grad_norm": 0.5721622109413147, |
| "learning_rate": 0.0003783406960456847, |
| "loss": 3.4664, |
| "step": 34400 |
| }, |
| { |
| "epoch": 3.7078893552900656, |
| "grad_norm": 0.626194953918457, |
| "learning_rate": 0.00037801745501562324, |
| "loss": 3.4689, |
| "step": 34450 |
| }, |
| { |
| "epoch": 3.713270907329674, |
| "grad_norm": 0.6698977947235107, |
| "learning_rate": 0.0003776942139855619, |
| "loss": 3.4617, |
| "step": 34500 |
| }, |
| { |
| "epoch": 3.7186524593692822, |
| "grad_norm": 0.5798797607421875, |
| "learning_rate": 0.0003773709729555005, |
| "loss": 3.4756, |
| "step": 34550 |
| }, |
| { |
| "epoch": 3.7240340114088903, |
| "grad_norm": 0.648026704788208, |
| "learning_rate": 0.000377047731925439, |
| "loss": 3.4632, |
| "step": 34600 |
| }, |
| { |
| "epoch": 3.7294155634484984, |
| "grad_norm": 0.6278716921806335, |
| "learning_rate": 0.0003767244908953776, |
| "loss": 3.4511, |
| "step": 34650 |
| }, |
| { |
| "epoch": 3.7347971154881066, |
| "grad_norm": 0.5706120729446411, |
| "learning_rate": 0.0003764012498653162, |
| "loss": 3.4438, |
| "step": 34700 |
| }, |
| { |
| "epoch": 3.740178667527715, |
| "grad_norm": 0.6078208088874817, |
| "learning_rate": 0.00037607800883525475, |
| "loss": 3.4435, |
| "step": 34750 |
| }, |
| { |
| "epoch": 3.745560219567323, |
| "grad_norm": 0.5775535702705383, |
| "learning_rate": 0.0003757547678051934, |
| "loss": 3.4463, |
| "step": 34800 |
| }, |
| { |
| "epoch": 3.7509417716069313, |
| "grad_norm": 0.6093581914901733, |
| "learning_rate": 0.000375431526775132, |
| "loss": 3.4498, |
| "step": 34850 |
| }, |
| { |
| "epoch": 3.75632332364654, |
| "grad_norm": 0.6318726539611816, |
| "learning_rate": 0.00037510828574507054, |
| "loss": 3.4469, |
| "step": 34900 |
| }, |
| { |
| "epoch": 3.761704875686148, |
| "grad_norm": 0.5926064252853394, |
| "learning_rate": 0.00037479150953561034, |
| "loss": 3.4331, |
| "step": 34950 |
| }, |
| { |
| "epoch": 3.767086427725756, |
| "grad_norm": 0.5718520283699036, |
| "learning_rate": 0.00037446826850554894, |
| "loss": 3.4391, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.767086427725756, |
| "eval_accuracy": 0.3750746581155395, |
| "eval_loss": 3.45552396774292, |
| "eval_runtime": 180.4933, |
| "eval_samples_per_second": 99.788, |
| "eval_steps_per_second": 6.238, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.772467979765364, |
| "grad_norm": 0.6313642859458923, |
| "learning_rate": 0.00037414502747548753, |
| "loss": 3.4635, |
| "step": 35050 |
| }, |
| { |
| "epoch": 3.7778495318049723, |
| "grad_norm": 0.6304598450660706, |
| "learning_rate": 0.00037382178644542607, |
| "loss": 3.4657, |
| "step": 35100 |
| }, |
| { |
| "epoch": 3.783231083844581, |
| "grad_norm": 0.5782265067100525, |
| "learning_rate": 0.00037349854541536467, |
| "loss": 3.4483, |
| "step": 35150 |
| }, |
| { |
| "epoch": 3.788612635884189, |
| "grad_norm": 0.6751189231872559, |
| "learning_rate": 0.0003731753043853033, |
| "loss": 3.4494, |
| "step": 35200 |
| }, |
| { |
| "epoch": 3.793994187923797, |
| "grad_norm": 0.6328338980674744, |
| "learning_rate": 0.00037285206335524186, |
| "loss": 3.4462, |
| "step": 35250 |
| }, |
| { |
| "epoch": 3.7993757399634056, |
| "grad_norm": 0.608933687210083, |
| "learning_rate": 0.00037252882232518045, |
| "loss": 3.451, |
| "step": 35300 |
| }, |
| { |
| "epoch": 3.8047572920030137, |
| "grad_norm": 0.5944675803184509, |
| "learning_rate": 0.000372205581295119, |
| "loss": 3.4391, |
| "step": 35350 |
| }, |
| { |
| "epoch": 3.810138844042622, |
| "grad_norm": 0.5831218957901001, |
| "learning_rate": 0.0003718823402650576, |
| "loss": 3.4609, |
| "step": 35400 |
| }, |
| { |
| "epoch": 3.8155203960822304, |
| "grad_norm": 0.6026729345321655, |
| "learning_rate": 0.00037155909923499624, |
| "loss": 3.43, |
| "step": 35450 |
| }, |
| { |
| "epoch": 3.8209019481218385, |
| "grad_norm": 0.5765225887298584, |
| "learning_rate": 0.0003712358582049348, |
| "loss": 3.4546, |
| "step": 35500 |
| }, |
| { |
| "epoch": 3.8262835001614466, |
| "grad_norm": 0.6434342861175537, |
| "learning_rate": 0.00037091261717487337, |
| "loss": 3.4432, |
| "step": 35550 |
| }, |
| { |
| "epoch": 3.8316650522010547, |
| "grad_norm": 0.6166746020317078, |
| "learning_rate": 0.00037058937614481197, |
| "loss": 3.4316, |
| "step": 35600 |
| }, |
| { |
| "epoch": 3.837046604240663, |
| "grad_norm": 0.5961298942565918, |
| "learning_rate": 0.0003702661351147505, |
| "loss": 3.449, |
| "step": 35650 |
| }, |
| { |
| "epoch": 3.8424281562802713, |
| "grad_norm": 0.5844510793685913, |
| "learning_rate": 0.0003699428940846891, |
| "loss": 3.4387, |
| "step": 35700 |
| }, |
| { |
| "epoch": 3.8478097083198795, |
| "grad_norm": 0.6083241105079651, |
| "learning_rate": 0.00036961965305462775, |
| "loss": 3.4513, |
| "step": 35750 |
| }, |
| { |
| "epoch": 3.8531912603594876, |
| "grad_norm": 0.5536696314811707, |
| "learning_rate": 0.0003692964120245663, |
| "loss": 3.435, |
| "step": 35800 |
| }, |
| { |
| "epoch": 3.858572812399096, |
| "grad_norm": 0.5852833390235901, |
| "learning_rate": 0.0003689731709945049, |
| "loss": 3.4571, |
| "step": 35850 |
| }, |
| { |
| "epoch": 3.863954364438704, |
| "grad_norm": 0.6565154194831848, |
| "learning_rate": 0.0003686499299644434, |
| "loss": 3.4436, |
| "step": 35900 |
| }, |
| { |
| "epoch": 3.8693359164783123, |
| "grad_norm": 0.6170244812965393, |
| "learning_rate": 0.000368326688934382, |
| "loss": 3.4601, |
| "step": 35950 |
| }, |
| { |
| "epoch": 3.8747174685179204, |
| "grad_norm": 0.6194032430648804, |
| "learning_rate": 0.0003680034479043206, |
| "loss": 3.4502, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.8747174685179204, |
| "eval_accuracy": 0.3757220120211382, |
| "eval_loss": 3.4532008171081543, |
| "eval_runtime": 180.0191, |
| "eval_samples_per_second": 100.05, |
| "eval_steps_per_second": 6.255, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.8800990205575285, |
| "grad_norm": 0.6787853837013245, |
| "learning_rate": 0.0003676802068742592, |
| "loss": 3.4645, |
| "step": 36050 |
| }, |
| { |
| "epoch": 3.885480572597137, |
| "grad_norm": 0.5699832439422607, |
| "learning_rate": 0.0003673569658441978, |
| "loss": 3.4565, |
| "step": 36100 |
| }, |
| { |
| "epoch": 3.890862124636745, |
| "grad_norm": 0.6348679661750793, |
| "learning_rate": 0.0003670337248141364, |
| "loss": 3.4306, |
| "step": 36150 |
| }, |
| { |
| "epoch": 3.8962436766763533, |
| "grad_norm": 0.611460268497467, |
| "learning_rate": 0.00036671048378407494, |
| "loss": 3.4289, |
| "step": 36200 |
| }, |
| { |
| "epoch": 3.901625228715962, |
| "grad_norm": 0.6488123536109924, |
| "learning_rate": 0.00036638724275401353, |
| "loss": 3.4341, |
| "step": 36250 |
| }, |
| { |
| "epoch": 3.90700678075557, |
| "grad_norm": 0.5952411890029907, |
| "learning_rate": 0.0003660640017239522, |
| "loss": 3.4465, |
| "step": 36300 |
| }, |
| { |
| "epoch": 3.912388332795178, |
| "grad_norm": 0.5958573222160339, |
| "learning_rate": 0.0003657407606938907, |
| "loss": 3.4567, |
| "step": 36350 |
| }, |
| { |
| "epoch": 3.9177698848347866, |
| "grad_norm": 0.6205569505691528, |
| "learning_rate": 0.0003654175196638293, |
| "loss": 3.4473, |
| "step": 36400 |
| }, |
| { |
| "epoch": 3.9231514368743947, |
| "grad_norm": 0.6399919390678406, |
| "learning_rate": 0.00036509427863376786, |
| "loss": 3.4574, |
| "step": 36450 |
| }, |
| { |
| "epoch": 3.928532988914003, |
| "grad_norm": 0.6450821161270142, |
| "learning_rate": 0.00036477103760370645, |
| "loss": 3.4467, |
| "step": 36500 |
| }, |
| { |
| "epoch": 3.933914540953611, |
| "grad_norm": 0.6391284465789795, |
| "learning_rate": 0.00036444779657364505, |
| "loss": 3.4279, |
| "step": 36550 |
| }, |
| { |
| "epoch": 3.939296092993219, |
| "grad_norm": 0.5766438245773315, |
| "learning_rate": 0.00036412455554358364, |
| "loss": 3.4366, |
| "step": 36600 |
| }, |
| { |
| "epoch": 3.9446776450328276, |
| "grad_norm": 0.6284584403038025, |
| "learning_rate": 0.00036380131451352224, |
| "loss": 3.4329, |
| "step": 36650 |
| }, |
| { |
| "epoch": 3.9500591970724357, |
| "grad_norm": 0.5596027374267578, |
| "learning_rate": 0.00036347807348346083, |
| "loss": 3.4525, |
| "step": 36700 |
| }, |
| { |
| "epoch": 3.955440749112044, |
| "grad_norm": 0.6084979772567749, |
| "learning_rate": 0.00036315483245339937, |
| "loss": 3.4452, |
| "step": 36750 |
| }, |
| { |
| "epoch": 3.9608223011516523, |
| "grad_norm": 0.6172415614128113, |
| "learning_rate": 0.00036283159142333797, |
| "loss": 3.4485, |
| "step": 36800 |
| }, |
| { |
| "epoch": 3.9662038531912605, |
| "grad_norm": 0.6101384162902832, |
| "learning_rate": 0.0003625083503932765, |
| "loss": 3.4566, |
| "step": 36850 |
| }, |
| { |
| "epoch": 3.9715854052308686, |
| "grad_norm": 0.5563173890113831, |
| "learning_rate": 0.00036218510936321516, |
| "loss": 3.4444, |
| "step": 36900 |
| }, |
| { |
| "epoch": 3.9769669572704767, |
| "grad_norm": 0.6216126084327698, |
| "learning_rate": 0.00036186186833315375, |
| "loss": 3.4486, |
| "step": 36950 |
| }, |
| { |
| "epoch": 3.9823485093100848, |
| "grad_norm": 0.6059814691543579, |
| "learning_rate": 0.0003615386273030923, |
| "loss": 3.4418, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.9823485093100848, |
| "eval_accuracy": 0.37635067763003277, |
| "eval_loss": 3.4460902214050293, |
| "eval_runtime": 180.6452, |
| "eval_samples_per_second": 99.704, |
| "eval_steps_per_second": 6.233, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.9877300613496933, |
| "grad_norm": 0.5866619944572449, |
| "learning_rate": 0.0003612153862730309, |
| "loss": 3.4566, |
| "step": 37050 |
| }, |
| { |
| "epoch": 3.9931116133893014, |
| "grad_norm": 0.6042488813400269, |
| "learning_rate": 0.0003608986100635707, |
| "loss": 3.4515, |
| "step": 37100 |
| }, |
| { |
| "epoch": 3.9984931654289095, |
| "grad_norm": 0.6132956147193909, |
| "learning_rate": 0.0003605753690335093, |
| "loss": 3.4339, |
| "step": 37150 |
| }, |
| { |
| "epoch": 4.003874717468518, |
| "grad_norm": 0.5894633531570435, |
| "learning_rate": 0.0003602521280034478, |
| "loss": 3.3714, |
| "step": 37200 |
| }, |
| { |
| "epoch": 4.009256269508126, |
| "grad_norm": 0.5937063694000244, |
| "learning_rate": 0.0003599288869733865, |
| "loss": 3.3621, |
| "step": 37250 |
| }, |
| { |
| "epoch": 4.014637821547734, |
| "grad_norm": 0.6420974135398865, |
| "learning_rate": 0.00035960564594332507, |
| "loss": 3.3542, |
| "step": 37300 |
| }, |
| { |
| "epoch": 4.020019373587343, |
| "grad_norm": 0.5973079800605774, |
| "learning_rate": 0.0003592824049132636, |
| "loss": 3.3511, |
| "step": 37350 |
| }, |
| { |
| "epoch": 4.0254009256269505, |
| "grad_norm": 0.5899896025657654, |
| "learning_rate": 0.0003589591638832022, |
| "loss": 3.3511, |
| "step": 37400 |
| }, |
| { |
| "epoch": 4.030782477666559, |
| "grad_norm": 0.6211423277854919, |
| "learning_rate": 0.0003586359228531408, |
| "loss": 3.365, |
| "step": 37450 |
| }, |
| { |
| "epoch": 4.036164029706168, |
| "grad_norm": 0.5809051394462585, |
| "learning_rate": 0.00035831268182307934, |
| "loss": 3.3682, |
| "step": 37500 |
| }, |
| { |
| "epoch": 4.041545581745775, |
| "grad_norm": 0.6257584691047668, |
| "learning_rate": 0.000357989440793018, |
| "loss": 3.3659, |
| "step": 37550 |
| }, |
| { |
| "epoch": 4.046927133785384, |
| "grad_norm": 0.6464672088623047, |
| "learning_rate": 0.0003576661997629566, |
| "loss": 3.3591, |
| "step": 37600 |
| }, |
| { |
| "epoch": 4.0523086858249915, |
| "grad_norm": 0.6366177201271057, |
| "learning_rate": 0.0003573429587328951, |
| "loss": 3.3684, |
| "step": 37650 |
| }, |
| { |
| "epoch": 4.0576902378646, |
| "grad_norm": 0.6766300797462463, |
| "learning_rate": 0.0003570197177028337, |
| "loss": 3.3451, |
| "step": 37700 |
| }, |
| { |
| "epoch": 4.063071789904209, |
| "grad_norm": 0.620211124420166, |
| "learning_rate": 0.00035669647667277226, |
| "loss": 3.3523, |
| "step": 37750 |
| }, |
| { |
| "epoch": 4.068453341943816, |
| "grad_norm": 0.6328372955322266, |
| "learning_rate": 0.00035637323564271085, |
| "loss": 3.369, |
| "step": 37800 |
| }, |
| { |
| "epoch": 4.073834893983425, |
| "grad_norm": 0.6072499752044678, |
| "learning_rate": 0.0003560499946126495, |
| "loss": 3.3667, |
| "step": 37850 |
| }, |
| { |
| "epoch": 4.079216446023033, |
| "grad_norm": 0.5885031819343567, |
| "learning_rate": 0.00035572675358258804, |
| "loss": 3.3755, |
| "step": 37900 |
| }, |
| { |
| "epoch": 4.084597998062641, |
| "grad_norm": 0.5657071471214294, |
| "learning_rate": 0.00035540351255252664, |
| "loss": 3.3716, |
| "step": 37950 |
| }, |
| { |
| "epoch": 4.08997955010225, |
| "grad_norm": 0.6271093487739563, |
| "learning_rate": 0.00035508027152246523, |
| "loss": 3.3718, |
| "step": 38000 |
| }, |
| { |
| "epoch": 4.08997955010225, |
| "eval_accuracy": 0.3768978535730665, |
| "eval_loss": 3.4471983909606934, |
| "eval_runtime": 180.3248, |
| "eval_samples_per_second": 99.881, |
| "eval_steps_per_second": 6.244, |
| "step": 38000 |
| }, |
| { |
| "epoch": 4.095361102141858, |
| "grad_norm": 0.6071150302886963, |
| "learning_rate": 0.0003547570304924038, |
| "loss": 3.3633, |
| "step": 38050 |
| }, |
| { |
| "epoch": 4.100742654181466, |
| "grad_norm": 0.5937423706054688, |
| "learning_rate": 0.0003544337894623424, |
| "loss": 3.3846, |
| "step": 38100 |
| }, |
| { |
| "epoch": 4.106124206221074, |
| "grad_norm": 0.5915195345878601, |
| "learning_rate": 0.000354110548432281, |
| "loss": 3.3781, |
| "step": 38150 |
| }, |
| { |
| "epoch": 4.111505758260682, |
| "grad_norm": 0.6321580410003662, |
| "learning_rate": 0.00035378730740221956, |
| "loss": 3.3696, |
| "step": 38200 |
| }, |
| { |
| "epoch": 4.1168873103002905, |
| "grad_norm": 0.6198523640632629, |
| "learning_rate": 0.00035346406637215815, |
| "loss": 3.3694, |
| "step": 38250 |
| }, |
| { |
| "epoch": 4.122268862339899, |
| "grad_norm": 0.6256786584854126, |
| "learning_rate": 0.0003531408253420967, |
| "loss": 3.375, |
| "step": 38300 |
| }, |
| { |
| "epoch": 4.127650414379507, |
| "grad_norm": 0.6233739256858826, |
| "learning_rate": 0.0003528175843120353, |
| "loss": 3.3891, |
| "step": 38350 |
| }, |
| { |
| "epoch": 4.133031966419115, |
| "grad_norm": 0.6164677143096924, |
| "learning_rate": 0.00035249434328197394, |
| "loss": 3.366, |
| "step": 38400 |
| }, |
| { |
| "epoch": 4.138413518458724, |
| "grad_norm": 0.6025086045265198, |
| "learning_rate": 0.0003521711022519125, |
| "loss": 3.3528, |
| "step": 38450 |
| }, |
| { |
| "epoch": 4.1437950704983315, |
| "grad_norm": 0.6623359322547913, |
| "learning_rate": 0.00035184786122185107, |
| "loss": 3.3703, |
| "step": 38500 |
| }, |
| { |
| "epoch": 4.14917662253794, |
| "grad_norm": 0.6005052924156189, |
| "learning_rate": 0.00035152462019178967, |
| "loss": 3.3746, |
| "step": 38550 |
| }, |
| { |
| "epoch": 4.154558174577549, |
| "grad_norm": 0.6520372033119202, |
| "learning_rate": 0.0003512013791617282, |
| "loss": 3.3717, |
| "step": 38600 |
| }, |
| { |
| "epoch": 4.159939726617156, |
| "grad_norm": 0.5965567827224731, |
| "learning_rate": 0.0003508781381316668, |
| "loss": 3.3652, |
| "step": 38650 |
| }, |
| { |
| "epoch": 4.165321278656765, |
| "grad_norm": 0.6479796171188354, |
| "learning_rate": 0.00035055489710160545, |
| "loss": 3.3732, |
| "step": 38700 |
| }, |
| { |
| "epoch": 4.1707028306963725, |
| "grad_norm": 0.5954575538635254, |
| "learning_rate": 0.000350231656071544, |
| "loss": 3.3845, |
| "step": 38750 |
| }, |
| { |
| "epoch": 4.176084382735981, |
| "grad_norm": 0.6750271916389465, |
| "learning_rate": 0.0003499084150414826, |
| "loss": 3.3609, |
| "step": 38800 |
| }, |
| { |
| "epoch": 4.18146593477559, |
| "grad_norm": 0.6075855493545532, |
| "learning_rate": 0.0003495851740114211, |
| "loss": 3.3825, |
| "step": 38850 |
| }, |
| { |
| "epoch": 4.186847486815197, |
| "grad_norm": 0.6466742753982544, |
| "learning_rate": 0.0003492619329813597, |
| "loss": 3.3572, |
| "step": 38900 |
| }, |
| { |
| "epoch": 4.192229038854806, |
| "grad_norm": 0.6090953350067139, |
| "learning_rate": 0.0003489386919512983, |
| "loss": 3.3637, |
| "step": 38950 |
| }, |
| { |
| "epoch": 4.197610590894414, |
| "grad_norm": 0.6194248795509338, |
| "learning_rate": 0.0003486154509212369, |
| "loss": 3.3859, |
| "step": 39000 |
| }, |
| { |
| "epoch": 4.197610590894414, |
| "eval_accuracy": 0.3771731799907884, |
| "eval_loss": 3.4430482387542725, |
| "eval_runtime": 180.581, |
| "eval_samples_per_second": 99.739, |
| "eval_steps_per_second": 6.235, |
| "step": 39000 |
| }, |
| { |
| "epoch": 4.202992142934022, |
| "grad_norm": 0.6370955109596252, |
| "learning_rate": 0.0003482922098911755, |
| "loss": 3.3882, |
| "step": 39050 |
| }, |
| { |
| "epoch": 4.208373694973631, |
| "grad_norm": 0.6059002876281738, |
| "learning_rate": 0.0003479689688611141, |
| "loss": 3.3683, |
| "step": 39100 |
| }, |
| { |
| "epoch": 4.213755247013238, |
| "grad_norm": 0.6166430711746216, |
| "learning_rate": 0.00034764572783105264, |
| "loss": 3.3749, |
| "step": 39150 |
| }, |
| { |
| "epoch": 4.219136799052847, |
| "grad_norm": 0.6343166828155518, |
| "learning_rate": 0.00034732248680099123, |
| "loss": 3.3703, |
| "step": 39200 |
| }, |
| { |
| "epoch": 4.224518351092455, |
| "grad_norm": 0.6673994064331055, |
| "learning_rate": 0.00034700571059153104, |
| "loss": 3.3923, |
| "step": 39250 |
| }, |
| { |
| "epoch": 4.229899903132063, |
| "grad_norm": 0.5913220643997192, |
| "learning_rate": 0.00034668246956146963, |
| "loss": 3.394, |
| "step": 39300 |
| }, |
| { |
| "epoch": 4.2352814551716715, |
| "grad_norm": 0.6904832720756531, |
| "learning_rate": 0.00034635922853140823, |
| "loss": 3.3711, |
| "step": 39350 |
| }, |
| { |
| "epoch": 4.24066300721128, |
| "grad_norm": 0.6442829966545105, |
| "learning_rate": 0.0003460359875013468, |
| "loss": 3.3899, |
| "step": 39400 |
| }, |
| { |
| "epoch": 4.246044559250888, |
| "grad_norm": 0.5766376852989197, |
| "learning_rate": 0.0003457127464712854, |
| "loss": 3.3693, |
| "step": 39450 |
| }, |
| { |
| "epoch": 4.251426111290496, |
| "grad_norm": 0.6282715797424316, |
| "learning_rate": 0.00034538950544122396, |
| "loss": 3.3743, |
| "step": 39500 |
| }, |
| { |
| "epoch": 4.256807663330104, |
| "grad_norm": 0.5965016484260559, |
| "learning_rate": 0.00034506626441116255, |
| "loss": 3.3786, |
| "step": 39550 |
| }, |
| { |
| "epoch": 4.2621892153697125, |
| "grad_norm": 0.6471324563026428, |
| "learning_rate": 0.0003447430233811011, |
| "loss": 3.3732, |
| "step": 39600 |
| }, |
| { |
| "epoch": 4.267570767409321, |
| "grad_norm": 0.6036005616188049, |
| "learning_rate": 0.00034441978235103974, |
| "loss": 3.3769, |
| "step": 39650 |
| }, |
| { |
| "epoch": 4.272952319448929, |
| "grad_norm": 0.6378051042556763, |
| "learning_rate": 0.00034409654132097834, |
| "loss": 3.3983, |
| "step": 39700 |
| }, |
| { |
| "epoch": 4.278333871488537, |
| "grad_norm": 0.6300690770149231, |
| "learning_rate": 0.0003437733002909169, |
| "loss": 3.3651, |
| "step": 39750 |
| }, |
| { |
| "epoch": 4.283715423528146, |
| "grad_norm": 0.5953393578529358, |
| "learning_rate": 0.00034345005926085547, |
| "loss": 3.3929, |
| "step": 39800 |
| }, |
| { |
| "epoch": 4.2890969755677535, |
| "grad_norm": 0.5986993312835693, |
| "learning_rate": 0.00034312681823079407, |
| "loss": 3.3837, |
| "step": 39850 |
| }, |
| { |
| "epoch": 4.294478527607362, |
| "grad_norm": 0.6578076481819153, |
| "learning_rate": 0.00034280357720073266, |
| "loss": 3.3791, |
| "step": 39900 |
| }, |
| { |
| "epoch": 4.299860079646971, |
| "grad_norm": 0.6022131443023682, |
| "learning_rate": 0.00034248033617067126, |
| "loss": 3.3983, |
| "step": 39950 |
| }, |
| { |
| "epoch": 4.305241631686578, |
| "grad_norm": 0.6208112835884094, |
| "learning_rate": 0.00034215709514060985, |
| "loss": 3.3809, |
| "step": 40000 |
| }, |
| { |
| "epoch": 4.305241631686578, |
| "eval_accuracy": 0.37746382646569215, |
| "eval_loss": 3.43764591217041, |
| "eval_runtime": 180.2919, |
| "eval_samples_per_second": 99.899, |
| "eval_steps_per_second": 6.245, |
| "step": 40000 |
| }, |
| { |
| "epoch": 4.310623183726187, |
| "grad_norm": 0.5868871212005615, |
| "learning_rate": 0.0003418338541105484, |
| "loss": 3.3803, |
| "step": 40050 |
| }, |
| { |
| "epoch": 4.3160047357657945, |
| "grad_norm": 0.5954568982124329, |
| "learning_rate": 0.000341510613080487, |
| "loss": 3.3836, |
| "step": 40100 |
| }, |
| { |
| "epoch": 4.321386287805403, |
| "grad_norm": 0.6300106644630432, |
| "learning_rate": 0.0003411873720504255, |
| "loss": 3.3725, |
| "step": 40150 |
| }, |
| { |
| "epoch": 4.326767839845012, |
| "grad_norm": 0.6355574131011963, |
| "learning_rate": 0.0003408641310203642, |
| "loss": 3.3852, |
| "step": 40200 |
| }, |
| { |
| "epoch": 4.332149391884619, |
| "grad_norm": 0.6317576766014099, |
| "learning_rate": 0.00034054088999030277, |
| "loss": 3.3944, |
| "step": 40250 |
| }, |
| { |
| "epoch": 4.337530943924228, |
| "grad_norm": 0.6220029592514038, |
| "learning_rate": 0.0003402176489602413, |
| "loss": 3.3843, |
| "step": 40300 |
| }, |
| { |
| "epoch": 4.342912495963836, |
| "grad_norm": 0.6280874013900757, |
| "learning_rate": 0.0003398944079301799, |
| "loss": 3.3892, |
| "step": 40350 |
| }, |
| { |
| "epoch": 4.348294048003444, |
| "grad_norm": 0.6555618047714233, |
| "learning_rate": 0.0003395711669001185, |
| "loss": 3.3741, |
| "step": 40400 |
| }, |
| { |
| "epoch": 4.3536756000430525, |
| "grad_norm": 0.6413003206253052, |
| "learning_rate": 0.00033924792587005704, |
| "loss": 3.3906, |
| "step": 40450 |
| }, |
| { |
| "epoch": 4.359057152082661, |
| "grad_norm": 0.6359059810638428, |
| "learning_rate": 0.0003389246848399957, |
| "loss": 3.3967, |
| "step": 40500 |
| }, |
| { |
| "epoch": 4.364438704122269, |
| "grad_norm": 0.6359884738922119, |
| "learning_rate": 0.0003386014438099343, |
| "loss": 3.3777, |
| "step": 40550 |
| }, |
| { |
| "epoch": 4.369820256161877, |
| "grad_norm": 0.6541460156440735, |
| "learning_rate": 0.0003382782027798728, |
| "loss": 3.3911, |
| "step": 40600 |
| }, |
| { |
| "epoch": 4.375201808201485, |
| "grad_norm": 0.6076342463493347, |
| "learning_rate": 0.0003379549617498114, |
| "loss": 3.388, |
| "step": 40650 |
| }, |
| { |
| "epoch": 4.3805833602410935, |
| "grad_norm": 0.6610375046730042, |
| "learning_rate": 0.00033763172071974996, |
| "loss": 3.3873, |
| "step": 40700 |
| }, |
| { |
| "epoch": 4.385964912280702, |
| "grad_norm": 0.6042363047599792, |
| "learning_rate": 0.00033730847968968855, |
| "loss": 3.3692, |
| "step": 40750 |
| }, |
| { |
| "epoch": 4.39134646432031, |
| "grad_norm": 0.6662680506706238, |
| "learning_rate": 0.0003369852386596272, |
| "loss": 3.3931, |
| "step": 40800 |
| }, |
| { |
| "epoch": 4.396728016359918, |
| "grad_norm": 0.6107580065727234, |
| "learning_rate": 0.00033666199762956574, |
| "loss": 3.3989, |
| "step": 40850 |
| }, |
| { |
| "epoch": 4.402109568399527, |
| "grad_norm": 0.6042559146881104, |
| "learning_rate": 0.00033633875659950434, |
| "loss": 3.3916, |
| "step": 40900 |
| }, |
| { |
| "epoch": 4.4074911204391345, |
| "grad_norm": 0.6182630062103271, |
| "learning_rate": 0.0003360155155694429, |
| "loss": 3.4034, |
| "step": 40950 |
| }, |
| { |
| "epoch": 4.412872672478743, |
| "grad_norm": 0.6265887022018433, |
| "learning_rate": 0.0003356922745393815, |
| "loss": 3.384, |
| "step": 41000 |
| }, |
| { |
| "epoch": 4.412872672478743, |
| "eval_accuracy": 0.3774822974566206, |
| "eval_loss": 3.438303232192993, |
| "eval_runtime": 180.488, |
| "eval_samples_per_second": 99.791, |
| "eval_steps_per_second": 6.239, |
| "step": 41000 |
| }, |
| { |
| "epoch": 4.418254224518351, |
| "grad_norm": 0.607819676399231, |
| "learning_rate": 0.0003353690335093201, |
| "loss": 3.3809, |
| "step": 41050 |
| }, |
| { |
| "epoch": 4.423635776557959, |
| "grad_norm": 0.6153405904769897, |
| "learning_rate": 0.0003350457924792587, |
| "loss": 3.3748, |
| "step": 41100 |
| }, |
| { |
| "epoch": 4.429017328597568, |
| "grad_norm": 0.6520572900772095, |
| "learning_rate": 0.00033472255144919726, |
| "loss": 3.3901, |
| "step": 41150 |
| }, |
| { |
| "epoch": 4.4343988806371755, |
| "grad_norm": 0.6429771780967712, |
| "learning_rate": 0.00033439931041913585, |
| "loss": 3.3964, |
| "step": 41200 |
| }, |
| { |
| "epoch": 4.439780432676784, |
| "grad_norm": 0.6534312963485718, |
| "learning_rate": 0.0003340760693890744, |
| "loss": 3.3768, |
| "step": 41250 |
| }, |
| { |
| "epoch": 4.445161984716393, |
| "grad_norm": 0.639164388179779, |
| "learning_rate": 0.00033375929317961425, |
| "loss": 3.3835, |
| "step": 41300 |
| }, |
| { |
| "epoch": 4.450543536756, |
| "grad_norm": 0.6154906153678894, |
| "learning_rate": 0.0003334360521495528, |
| "loss": 3.4067, |
| "step": 41350 |
| }, |
| { |
| "epoch": 4.455925088795609, |
| "grad_norm": 0.6143419146537781, |
| "learning_rate": 0.0003331128111194914, |
| "loss": 3.3849, |
| "step": 41400 |
| }, |
| { |
| "epoch": 4.461306640835216, |
| "grad_norm": 0.6307302117347717, |
| "learning_rate": 0.00033278957008943004, |
| "loss": 3.3831, |
| "step": 41450 |
| }, |
| { |
| "epoch": 4.466688192874825, |
| "grad_norm": 0.6415225267410278, |
| "learning_rate": 0.0003324663290593686, |
| "loss": 3.3719, |
| "step": 41500 |
| }, |
| { |
| "epoch": 4.4720697449144335, |
| "grad_norm": 0.594627320766449, |
| "learning_rate": 0.00033214308802930717, |
| "loss": 3.3902, |
| "step": 41550 |
| }, |
| { |
| "epoch": 4.477451296954041, |
| "grad_norm": 0.6894360184669495, |
| "learning_rate": 0.0003318198469992457, |
| "loss": 3.398, |
| "step": 41600 |
| }, |
| { |
| "epoch": 4.48283284899365, |
| "grad_norm": 0.5951968431472778, |
| "learning_rate": 0.0003314966059691843, |
| "loss": 3.3817, |
| "step": 41650 |
| }, |
| { |
| "epoch": 4.488214401033258, |
| "grad_norm": 0.6982078552246094, |
| "learning_rate": 0.00033117336493912296, |
| "loss": 3.3932, |
| "step": 41700 |
| }, |
| { |
| "epoch": 4.493595953072866, |
| "grad_norm": 0.6590989828109741, |
| "learning_rate": 0.0003308501239090615, |
| "loss": 3.3916, |
| "step": 41750 |
| }, |
| { |
| "epoch": 4.4989775051124745, |
| "grad_norm": 0.626447319984436, |
| "learning_rate": 0.0003305268828790001, |
| "loss": 3.3732, |
| "step": 41800 |
| }, |
| { |
| "epoch": 4.504359057152083, |
| "grad_norm": 0.624427080154419, |
| "learning_rate": 0.0003302036418489387, |
| "loss": 3.3799, |
| "step": 41850 |
| }, |
| { |
| "epoch": 4.509740609191691, |
| "grad_norm": 0.6644538640975952, |
| "learning_rate": 0.0003298804008188772, |
| "loss": 3.4, |
| "step": 41900 |
| }, |
| { |
| "epoch": 4.515122161231299, |
| "grad_norm": 0.6066876649856567, |
| "learning_rate": 0.0003295571597888158, |
| "loss": 3.3932, |
| "step": 41950 |
| }, |
| { |
| "epoch": 4.520503713270907, |
| "grad_norm": 0.5975305438041687, |
| "learning_rate": 0.00032923391875875447, |
| "loss": 3.3768, |
| "step": 42000 |
| }, |
| { |
| "epoch": 4.520503713270907, |
| "eval_accuracy": 0.37895345755762866, |
| "eval_loss": 3.4301979541778564, |
| "eval_runtime": 180.422, |
| "eval_samples_per_second": 99.827, |
| "eval_steps_per_second": 6.241, |
| "step": 42000 |
| }, |
| { |
| "epoch": 4.5258852653105155, |
| "grad_norm": 0.6146498918533325, |
| "learning_rate": 0.000328910677728693, |
| "loss": 3.3973, |
| "step": 42050 |
| }, |
| { |
| "epoch": 4.531266817350124, |
| "grad_norm": 0.6150146722793579, |
| "learning_rate": 0.0003285874366986316, |
| "loss": 3.3871, |
| "step": 42100 |
| }, |
| { |
| "epoch": 4.536648369389732, |
| "grad_norm": 0.5906317830085754, |
| "learning_rate": 0.00032826419566857015, |
| "loss": 3.3871, |
| "step": 42150 |
| }, |
| { |
| "epoch": 4.54202992142934, |
| "grad_norm": 0.6632665395736694, |
| "learning_rate": 0.00032794095463850874, |
| "loss": 3.3829, |
| "step": 42200 |
| }, |
| { |
| "epoch": 4.547411473468949, |
| "grad_norm": 0.6187365055084229, |
| "learning_rate": 0.0003276177136084473, |
| "loss": 3.3977, |
| "step": 42250 |
| }, |
| { |
| "epoch": 4.5527930255085565, |
| "grad_norm": 0.6101699471473694, |
| "learning_rate": 0.00032729447257838593, |
| "loss": 3.3902, |
| "step": 42300 |
| }, |
| { |
| "epoch": 4.558174577548165, |
| "grad_norm": 0.6219088435173035, |
| "learning_rate": 0.0003269712315483245, |
| "loss": 3.3671, |
| "step": 42350 |
| }, |
| { |
| "epoch": 4.563556129587774, |
| "grad_norm": 0.6886587142944336, |
| "learning_rate": 0.00032664799051826306, |
| "loss": 3.3877, |
| "step": 42400 |
| }, |
| { |
| "epoch": 4.568937681627381, |
| "grad_norm": 0.7076488137245178, |
| "learning_rate": 0.00032632474948820166, |
| "loss": 3.3835, |
| "step": 42450 |
| }, |
| { |
| "epoch": 4.57431923366699, |
| "grad_norm": 0.6349630355834961, |
| "learning_rate": 0.00032600150845814025, |
| "loss": 3.4027, |
| "step": 42500 |
| }, |
| { |
| "epoch": 4.579700785706597, |
| "grad_norm": 0.6757253408432007, |
| "learning_rate": 0.0003256782674280788, |
| "loss": 3.376, |
| "step": 42550 |
| }, |
| { |
| "epoch": 4.585082337746206, |
| "grad_norm": 0.6440753936767578, |
| "learning_rate": 0.00032535502639801744, |
| "loss": 3.387, |
| "step": 42600 |
| }, |
| { |
| "epoch": 4.5904638897858145, |
| "grad_norm": 0.5789088606834412, |
| "learning_rate": 0.00032503178536795604, |
| "loss": 3.3952, |
| "step": 42650 |
| }, |
| { |
| "epoch": 4.595845441825422, |
| "grad_norm": 0.6196984052658081, |
| "learning_rate": 0.0003247085443378946, |
| "loss": 3.3845, |
| "step": 42700 |
| }, |
| { |
| "epoch": 4.601226993865031, |
| "grad_norm": 0.5972916483879089, |
| "learning_rate": 0.0003243853033078332, |
| "loss": 3.3952, |
| "step": 42750 |
| }, |
| { |
| "epoch": 4.606608545904638, |
| "grad_norm": 0.6403440237045288, |
| "learning_rate": 0.0003240620622777717, |
| "loss": 3.3873, |
| "step": 42800 |
| }, |
| { |
| "epoch": 4.611990097944247, |
| "grad_norm": 0.6815773844718933, |
| "learning_rate": 0.00032373882124771036, |
| "loss": 3.415, |
| "step": 42850 |
| }, |
| { |
| "epoch": 4.6173716499838555, |
| "grad_norm": 0.6044899225234985, |
| "learning_rate": 0.00032341558021764896, |
| "loss": 3.385, |
| "step": 42900 |
| }, |
| { |
| "epoch": 4.622753202023463, |
| "grad_norm": 0.6453890800476074, |
| "learning_rate": 0.0003230923391875875, |
| "loss": 3.3874, |
| "step": 42950 |
| }, |
| { |
| "epoch": 4.628134754063072, |
| "grad_norm": 0.6140652298927307, |
| "learning_rate": 0.0003227690981575261, |
| "loss": 3.381, |
| "step": 43000 |
| }, |
| { |
| "epoch": 4.628134754063072, |
| "eval_accuracy": 0.3790061542082187, |
| "eval_loss": 3.4236814975738525, |
| "eval_runtime": 180.3345, |
| "eval_samples_per_second": 99.876, |
| "eval_steps_per_second": 6.244, |
| "step": 43000 |
| }, |
| { |
| "epoch": 4.63351630610268, |
| "grad_norm": 0.6114611029624939, |
| "learning_rate": 0.0003224458571274647, |
| "loss": 3.3835, |
| "step": 43050 |
| }, |
| { |
| "epoch": 4.638897858142288, |
| "grad_norm": 0.621130645275116, |
| "learning_rate": 0.00032212261609740323, |
| "loss": 3.3746, |
| "step": 43100 |
| }, |
| { |
| "epoch": 4.6442794101818965, |
| "grad_norm": 0.7222752571105957, |
| "learning_rate": 0.0003217993750673419, |
| "loss": 3.4018, |
| "step": 43150 |
| }, |
| { |
| "epoch": 4.649660962221505, |
| "grad_norm": 0.6595478057861328, |
| "learning_rate": 0.00032147613403728047, |
| "loss": 3.4018, |
| "step": 43200 |
| }, |
| { |
| "epoch": 4.655042514261113, |
| "grad_norm": 0.6354619860649109, |
| "learning_rate": 0.000321152893007219, |
| "loss": 3.403, |
| "step": 43250 |
| }, |
| { |
| "epoch": 4.660424066300721, |
| "grad_norm": 0.7083771228790283, |
| "learning_rate": 0.0003208296519771576, |
| "loss": 3.3735, |
| "step": 43300 |
| }, |
| { |
| "epoch": 4.665805618340329, |
| "grad_norm": 0.62490314245224, |
| "learning_rate": 0.0003205128757676974, |
| "loss": 3.3931, |
| "step": 43350 |
| }, |
| { |
| "epoch": 4.6711871703799375, |
| "grad_norm": 0.6417445540428162, |
| "learning_rate": 0.000320189634737636, |
| "loss": 3.402, |
| "step": 43400 |
| }, |
| { |
| "epoch": 4.676568722419546, |
| "grad_norm": 0.613502025604248, |
| "learning_rate": 0.00031986639370757455, |
| "loss": 3.3852, |
| "step": 43450 |
| }, |
| { |
| "epoch": 4.681950274459154, |
| "grad_norm": 0.5854348540306091, |
| "learning_rate": 0.0003195431526775132, |
| "loss": 3.3849, |
| "step": 43500 |
| }, |
| { |
| "epoch": 4.687331826498762, |
| "grad_norm": 0.6281751990318298, |
| "learning_rate": 0.0003192199116474518, |
| "loss": 3.3938, |
| "step": 43550 |
| }, |
| { |
| "epoch": 4.692713378538371, |
| "grad_norm": 0.6472344398498535, |
| "learning_rate": 0.00031889667061739033, |
| "loss": 3.3916, |
| "step": 43600 |
| }, |
| { |
| "epoch": 4.6980949305779784, |
| "grad_norm": 0.66495680809021, |
| "learning_rate": 0.0003185734295873289, |
| "loss": 3.3904, |
| "step": 43650 |
| }, |
| { |
| "epoch": 4.703476482617587, |
| "grad_norm": 0.6219391822814941, |
| "learning_rate": 0.00031825665337786873, |
| "loss": 3.3885, |
| "step": 43700 |
| }, |
| { |
| "epoch": 4.7088580346571955, |
| "grad_norm": 0.5999084115028381, |
| "learning_rate": 0.00031793341234780733, |
| "loss": 3.3887, |
| "step": 43750 |
| }, |
| { |
| "epoch": 4.714239586696803, |
| "grad_norm": 0.6449667811393738, |
| "learning_rate": 0.00031761017131774587, |
| "loss": 3.3878, |
| "step": 43800 |
| }, |
| { |
| "epoch": 4.719621138736412, |
| "grad_norm": 0.6235863566398621, |
| "learning_rate": 0.00031728693028768446, |
| "loss": 3.3608, |
| "step": 43850 |
| }, |
| { |
| "epoch": 4.725002690776019, |
| "grad_norm": 0.6473997831344604, |
| "learning_rate": 0.0003169636892576231, |
| "loss": 3.3974, |
| "step": 43900 |
| }, |
| { |
| "epoch": 4.730384242815628, |
| "grad_norm": 0.6401399970054626, |
| "learning_rate": 0.00031664044822756165, |
| "loss": 3.4097, |
| "step": 43950 |
| }, |
| { |
| "epoch": 4.7357657948552365, |
| "grad_norm": 0.6407394409179688, |
| "learning_rate": 0.00031631720719750025, |
| "loss": 3.3897, |
| "step": 44000 |
| }, |
| { |
| "epoch": 4.7357657948552365, |
| "eval_accuracy": 0.3795691934728733, |
| "eval_loss": 3.419826030731201, |
| "eval_runtime": 180.2132, |
| "eval_samples_per_second": 99.943, |
| "eval_steps_per_second": 6.248, |
| "step": 44000 |
| }, |
| { |
| "epoch": 4.741147346894844, |
| "grad_norm": 0.6270405650138855, |
| "learning_rate": 0.00031599396616743884, |
| "loss": 3.3685, |
| "step": 44050 |
| }, |
| { |
| "epoch": 4.746528898934453, |
| "grad_norm": 0.6910510063171387, |
| "learning_rate": 0.0003156707251373774, |
| "loss": 3.3945, |
| "step": 44100 |
| }, |
| { |
| "epoch": 4.751910450974061, |
| "grad_norm": 0.602331280708313, |
| "learning_rate": 0.000315347484107316, |
| "loss": 3.3932, |
| "step": 44150 |
| }, |
| { |
| "epoch": 4.757292003013669, |
| "grad_norm": 0.6058678030967712, |
| "learning_rate": 0.0003150242430772546, |
| "loss": 3.3845, |
| "step": 44200 |
| }, |
| { |
| "epoch": 4.7626735550532775, |
| "grad_norm": 0.6464012861251831, |
| "learning_rate": 0.00031470100204719317, |
| "loss": 3.3934, |
| "step": 44250 |
| }, |
| { |
| "epoch": 4.768055107092886, |
| "grad_norm": 0.6263419985771179, |
| "learning_rate": 0.00031437776101713176, |
| "loss": 3.3844, |
| "step": 44300 |
| }, |
| { |
| "epoch": 4.773436659132494, |
| "grad_norm": 0.6222793459892273, |
| "learning_rate": 0.0003140545199870703, |
| "loss": 3.385, |
| "step": 44350 |
| }, |
| { |
| "epoch": 4.778818211172102, |
| "grad_norm": 0.7262570261955261, |
| "learning_rate": 0.0003137312789570089, |
| "loss": 3.3795, |
| "step": 44400 |
| }, |
| { |
| "epoch": 4.78419976321171, |
| "grad_norm": 0.625326931476593, |
| "learning_rate": 0.00031340803792694754, |
| "loss": 3.3824, |
| "step": 44450 |
| }, |
| { |
| "epoch": 4.7895813152513185, |
| "grad_norm": 0.6770031452178955, |
| "learning_rate": 0.0003130847968968861, |
| "loss": 3.3799, |
| "step": 44500 |
| }, |
| { |
| "epoch": 4.794962867290927, |
| "grad_norm": 0.6451786160469055, |
| "learning_rate": 0.0003127615558668247, |
| "loss": 3.3916, |
| "step": 44550 |
| }, |
| { |
| "epoch": 4.800344419330535, |
| "grad_norm": 0.6783530116081238, |
| "learning_rate": 0.0003124383148367633, |
| "loss": 3.3967, |
| "step": 44600 |
| }, |
| { |
| "epoch": 4.805725971370143, |
| "grad_norm": 0.6572859287261963, |
| "learning_rate": 0.0003121150738067018, |
| "loss": 3.3786, |
| "step": 44650 |
| }, |
| { |
| "epoch": 4.811107523409751, |
| "grad_norm": 0.6286121606826782, |
| "learning_rate": 0.0003117918327766404, |
| "loss": 3.3847, |
| "step": 44700 |
| }, |
| { |
| "epoch": 4.8164890754493594, |
| "grad_norm": 0.6354442238807678, |
| "learning_rate": 0.00031146859174657906, |
| "loss": 3.3885, |
| "step": 44750 |
| }, |
| { |
| "epoch": 4.821870627488968, |
| "grad_norm": 0.6219363212585449, |
| "learning_rate": 0.0003111453507165176, |
| "loss": 3.3731, |
| "step": 44800 |
| }, |
| { |
| "epoch": 4.827252179528576, |
| "grad_norm": 0.6498527526855469, |
| "learning_rate": 0.0003108221096864562, |
| "loss": 3.3757, |
| "step": 44850 |
| }, |
| { |
| "epoch": 4.832633731568184, |
| "grad_norm": 0.6466692090034485, |
| "learning_rate": 0.00031049886865639473, |
| "loss": 3.3938, |
| "step": 44900 |
| }, |
| { |
| "epoch": 4.838015283607793, |
| "grad_norm": 0.6414273381233215, |
| "learning_rate": 0.00031017562762633333, |
| "loss": 3.3908, |
| "step": 44950 |
| }, |
| { |
| "epoch": 4.8433968356474, |
| "grad_norm": 0.6336743831634521, |
| "learning_rate": 0.00030985238659627187, |
| "loss": 3.4056, |
| "step": 45000 |
| }, |
| { |
| "epoch": 4.8433968356474, |
| "eval_accuracy": 0.37976868017490073, |
| "eval_loss": 3.4165797233581543, |
| "eval_runtime": 179.7792, |
| "eval_samples_per_second": 100.184, |
| "eval_steps_per_second": 6.263, |
| "step": 45000 |
| }, |
| { |
| "epoch": 4.848778387687009, |
| "grad_norm": 0.6573559045791626, |
| "learning_rate": 0.0003095291455662105, |
| "loss": 3.4003, |
| "step": 45050 |
| }, |
| { |
| "epoch": 4.8541599397266175, |
| "grad_norm": 0.6122683882713318, |
| "learning_rate": 0.0003092059045361491, |
| "loss": 3.4081, |
| "step": 45100 |
| }, |
| { |
| "epoch": 4.859541491766225, |
| "grad_norm": 0.6225839853286743, |
| "learning_rate": 0.00030888266350608765, |
| "loss": 3.3859, |
| "step": 45150 |
| }, |
| { |
| "epoch": 4.864923043805834, |
| "grad_norm": 0.6035463213920593, |
| "learning_rate": 0.00030855942247602625, |
| "loss": 3.3995, |
| "step": 45200 |
| }, |
| { |
| "epoch": 4.870304595845441, |
| "grad_norm": 0.620333194732666, |
| "learning_rate": 0.00030823618144596484, |
| "loss": 3.3876, |
| "step": 45250 |
| }, |
| { |
| "epoch": 4.87568614788505, |
| "grad_norm": 0.6101974844932556, |
| "learning_rate": 0.0003079129404159035, |
| "loss": 3.3992, |
| "step": 45300 |
| }, |
| { |
| "epoch": 4.8810676999246585, |
| "grad_norm": 0.6473581194877625, |
| "learning_rate": 0.00030758969938584203, |
| "loss": 3.4209, |
| "step": 45350 |
| }, |
| { |
| "epoch": 4.886449251964266, |
| "grad_norm": 0.6419587731361389, |
| "learning_rate": 0.0003072664583557806, |
| "loss": 3.3862, |
| "step": 45400 |
| }, |
| { |
| "epoch": 4.891830804003875, |
| "grad_norm": 0.6355916261672974, |
| "learning_rate": 0.00030694321732571917, |
| "loss": 3.3831, |
| "step": 45450 |
| }, |
| { |
| "epoch": 4.897212356043483, |
| "grad_norm": 0.6475318670272827, |
| "learning_rate": 0.00030661997629565776, |
| "loss": 3.3749, |
| "step": 45500 |
| }, |
| { |
| "epoch": 4.902593908083091, |
| "grad_norm": 0.6229646801948547, |
| "learning_rate": 0.0003062967352655963, |
| "loss": 3.3826, |
| "step": 45550 |
| }, |
| { |
| "epoch": 4.9079754601226995, |
| "grad_norm": 0.6335062384605408, |
| "learning_rate": 0.00030597349423553495, |
| "loss": 3.3829, |
| "step": 45600 |
| }, |
| { |
| "epoch": 4.913357012162308, |
| "grad_norm": 0.6275141835212708, |
| "learning_rate": 0.00030565025320547355, |
| "loss": 3.3979, |
| "step": 45650 |
| }, |
| { |
| "epoch": 4.918738564201916, |
| "grad_norm": 0.6986739039421082, |
| "learning_rate": 0.0003053270121754121, |
| "loss": 3.4014, |
| "step": 45700 |
| }, |
| { |
| "epoch": 4.924120116241524, |
| "grad_norm": 0.7142667174339294, |
| "learning_rate": 0.0003050037711453507, |
| "loss": 3.3784, |
| "step": 45750 |
| }, |
| { |
| "epoch": 4.929501668281132, |
| "grad_norm": 0.6080502271652222, |
| "learning_rate": 0.0003046805301152893, |
| "loss": 3.3835, |
| "step": 45800 |
| }, |
| { |
| "epoch": 4.9348832203207404, |
| "grad_norm": 0.6908614635467529, |
| "learning_rate": 0.0003043572890852278, |
| "loss": 3.3739, |
| "step": 45850 |
| }, |
| { |
| "epoch": 4.940264772360349, |
| "grad_norm": 0.653312623500824, |
| "learning_rate": 0.00030403404805516646, |
| "loss": 3.3974, |
| "step": 45900 |
| }, |
| { |
| "epoch": 4.945646324399957, |
| "grad_norm": 0.6278221607208252, |
| "learning_rate": 0.00030371080702510506, |
| "loss": 3.3654, |
| "step": 45950 |
| }, |
| { |
| "epoch": 4.951027876439565, |
| "grad_norm": 0.6530224680900574, |
| "learning_rate": 0.0003033875659950436, |
| "loss": 3.3793, |
| "step": 46000 |
| }, |
| { |
| "epoch": 4.951027876439565, |
| "eval_accuracy": 0.3805414195130373, |
| "eval_loss": 3.408653974533081, |
| "eval_runtime": 180.4298, |
| "eval_samples_per_second": 99.823, |
| "eval_steps_per_second": 6.241, |
| "step": 46000 |
| }, |
| { |
| "epoch": 4.956409428479174, |
| "grad_norm": 0.6106470227241516, |
| "learning_rate": 0.0003030643249649822, |
| "loss": 3.3896, |
| "step": 46050 |
| }, |
| { |
| "epoch": 4.961790980518781, |
| "grad_norm": 0.620707631111145, |
| "learning_rate": 0.00030274108393492073, |
| "loss": 3.3754, |
| "step": 46100 |
| }, |
| { |
| "epoch": 4.96717253255839, |
| "grad_norm": 0.6534816026687622, |
| "learning_rate": 0.00030241784290485933, |
| "loss": 3.3799, |
| "step": 46150 |
| }, |
| { |
| "epoch": 4.9725540845979985, |
| "grad_norm": 0.6243362426757812, |
| "learning_rate": 0.000302094601874798, |
| "loss": 3.3661, |
| "step": 46200 |
| }, |
| { |
| "epoch": 4.977935636637606, |
| "grad_norm": 0.6367123126983643, |
| "learning_rate": 0.0003017713608447365, |
| "loss": 3.3833, |
| "step": 46250 |
| }, |
| { |
| "epoch": 4.983317188677215, |
| "grad_norm": 0.6724542379379272, |
| "learning_rate": 0.0003014481198146751, |
| "loss": 3.3736, |
| "step": 46300 |
| }, |
| { |
| "epoch": 4.988698740716822, |
| "grad_norm": 0.6005682349205017, |
| "learning_rate": 0.0003011248787846137, |
| "loss": 3.3951, |
| "step": 46350 |
| }, |
| { |
| "epoch": 4.994080292756431, |
| "grad_norm": 0.6530224084854126, |
| "learning_rate": 0.00030080163775455225, |
| "loss": 3.3865, |
| "step": 46400 |
| }, |
| { |
| "epoch": 4.9994618447960395, |
| "grad_norm": 0.6524396538734436, |
| "learning_rate": 0.0003004783967244909, |
| "loss": 3.3978, |
| "step": 46450 |
| }, |
| { |
| "epoch": 5.004843396835647, |
| "grad_norm": 0.664138674736023, |
| "learning_rate": 0.0003001551556944295, |
| "loss": 3.2914, |
| "step": 46500 |
| }, |
| { |
| "epoch": 5.010224948875256, |
| "grad_norm": 0.6500561237335205, |
| "learning_rate": 0.00029983191466436803, |
| "loss": 3.2781, |
| "step": 46550 |
| }, |
| { |
| "epoch": 5.015606500914864, |
| "grad_norm": 0.6164722442626953, |
| "learning_rate": 0.0002995086736343066, |
| "loss": 3.2889, |
| "step": 46600 |
| }, |
| { |
| "epoch": 5.020988052954472, |
| "grad_norm": 0.6303194761276245, |
| "learning_rate": 0.00029918543260424517, |
| "loss": 3.2927, |
| "step": 46650 |
| }, |
| { |
| "epoch": 5.0263696049940805, |
| "grad_norm": 0.640316367149353, |
| "learning_rate": 0.0002988621915741838, |
| "loss": 3.3064, |
| "step": 46700 |
| }, |
| { |
| "epoch": 5.031751157033688, |
| "grad_norm": 0.6622575521469116, |
| "learning_rate": 0.00029853895054412236, |
| "loss": 3.3021, |
| "step": 46750 |
| }, |
| { |
| "epoch": 5.037132709073297, |
| "grad_norm": 0.5933485627174377, |
| "learning_rate": 0.00029821570951406095, |
| "loss": 3.294, |
| "step": 46800 |
| }, |
| { |
| "epoch": 5.042514261112905, |
| "grad_norm": 0.6456048488616943, |
| "learning_rate": 0.00029789246848399955, |
| "loss": 3.2845, |
| "step": 46850 |
| }, |
| { |
| "epoch": 5.047895813152513, |
| "grad_norm": 0.6657723188400269, |
| "learning_rate": 0.00029756922745393814, |
| "loss": 3.3037, |
| "step": 46900 |
| }, |
| { |
| "epoch": 5.0532773651921215, |
| "grad_norm": 0.6260988116264343, |
| "learning_rate": 0.00029724598642387674, |
| "loss": 3.3016, |
| "step": 46950 |
| }, |
| { |
| "epoch": 5.05865891723173, |
| "grad_norm": 0.6429811716079712, |
| "learning_rate": 0.0002969227453938153, |
| "loss": 3.294, |
| "step": 47000 |
| }, |
| { |
| "epoch": 5.05865891723173, |
| "eval_accuracy": 0.3809833108077788, |
| "eval_loss": 3.4148974418640137, |
| "eval_runtime": 180.1259, |
| "eval_samples_per_second": 99.991, |
| "eval_steps_per_second": 6.251, |
| "step": 47000 |
| }, |
| { |
| "epoch": 5.064040469271338, |
| "grad_norm": 0.6801566481590271, |
| "learning_rate": 0.00029659950436375387, |
| "loss": 3.3172, |
| "step": 47050 |
| }, |
| { |
| "epoch": 5.069422021310946, |
| "grad_norm": 0.636012852191925, |
| "learning_rate": 0.00029627626333369246, |
| "loss": 3.3021, |
| "step": 47100 |
| }, |
| { |
| "epoch": 5.074803573350554, |
| "grad_norm": 0.6584773063659668, |
| "learning_rate": 0.00029595302230363106, |
| "loss": 3.3106, |
| "step": 47150 |
| }, |
| { |
| "epoch": 5.080185125390162, |
| "grad_norm": 0.656566321849823, |
| "learning_rate": 0.0002956297812735696, |
| "loss": 3.3225, |
| "step": 47200 |
| }, |
| { |
| "epoch": 5.085566677429771, |
| "grad_norm": 0.6332939267158508, |
| "learning_rate": 0.00029530654024350825, |
| "loss": 3.3065, |
| "step": 47250 |
| }, |
| { |
| "epoch": 5.090948229469379, |
| "grad_norm": 0.6884245872497559, |
| "learning_rate": 0.0002949832992134468, |
| "loss": 3.2999, |
| "step": 47300 |
| }, |
| { |
| "epoch": 5.096329781508987, |
| "grad_norm": 0.6462180018424988, |
| "learning_rate": 0.0002946600581833854, |
| "loss": 3.3009, |
| "step": 47350 |
| }, |
| { |
| "epoch": 5.101711333548596, |
| "grad_norm": 0.6998089551925659, |
| "learning_rate": 0.000294336817153324, |
| "loss": 3.3156, |
| "step": 47400 |
| }, |
| { |
| "epoch": 5.107092885588203, |
| "grad_norm": 0.6399083733558655, |
| "learning_rate": 0.0002940135761232626, |
| "loss": 3.3166, |
| "step": 47450 |
| }, |
| { |
| "epoch": 5.112474437627812, |
| "grad_norm": 0.6904787421226501, |
| "learning_rate": 0.0002936903350932011, |
| "loss": 3.3293, |
| "step": 47500 |
| }, |
| { |
| "epoch": 5.1178559896674205, |
| "grad_norm": 0.7392804026603699, |
| "learning_rate": 0.0002933670940631397, |
| "loss": 3.3106, |
| "step": 47550 |
| }, |
| { |
| "epoch": 5.123237541707028, |
| "grad_norm": 0.6379286050796509, |
| "learning_rate": 0.0002930438530330783, |
| "loss": 3.3172, |
| "step": 47600 |
| }, |
| { |
| "epoch": 5.128619093746637, |
| "grad_norm": 0.6620005965232849, |
| "learning_rate": 0.0002927206120030169, |
| "loss": 3.3154, |
| "step": 47650 |
| }, |
| { |
| "epoch": 5.134000645786244, |
| "grad_norm": 0.6485711932182312, |
| "learning_rate": 0.0002923973709729555, |
| "loss": 3.3219, |
| "step": 47700 |
| }, |
| { |
| "epoch": 5.139382197825853, |
| "grad_norm": 0.6479270458221436, |
| "learning_rate": 0.00029207412994289403, |
| "loss": 3.3062, |
| "step": 47750 |
| }, |
| { |
| "epoch": 5.1447637498654615, |
| "grad_norm": 0.6357467770576477, |
| "learning_rate": 0.0002917508889128327, |
| "loss": 3.3221, |
| "step": 47800 |
| }, |
| { |
| "epoch": 5.150145301905069, |
| "grad_norm": 0.7255944013595581, |
| "learning_rate": 0.0002914276478827712, |
| "loss": 3.3055, |
| "step": 47850 |
| }, |
| { |
| "epoch": 5.155526853944678, |
| "grad_norm": 0.6277051568031311, |
| "learning_rate": 0.00029111087167331103, |
| "loss": 3.2959, |
| "step": 47900 |
| }, |
| { |
| "epoch": 5.160908405984286, |
| "grad_norm": 0.6564116477966309, |
| "learning_rate": 0.0002907876306432496, |
| "loss": 3.3198, |
| "step": 47950 |
| }, |
| { |
| "epoch": 5.166289958023894, |
| "grad_norm": 0.6674607396125793, |
| "learning_rate": 0.0002904643896131882, |
| "loss": 3.3209, |
| "step": 48000 |
| }, |
| { |
| "epoch": 5.166289958023894, |
| "eval_accuracy": 0.38105034963956036, |
| "eval_loss": 3.410404920578003, |
| "eval_runtime": 180.0365, |
| "eval_samples_per_second": 100.041, |
| "eval_steps_per_second": 6.254, |
| "step": 48000 |
| }, |
| { |
| "epoch": 5.1716715100635025, |
| "grad_norm": 0.6171758770942688, |
| "learning_rate": 0.0002901411485831268, |
| "loss": 3.319, |
| "step": 48050 |
| }, |
| { |
| "epoch": 5.17705306210311, |
| "grad_norm": 0.7050843238830566, |
| "learning_rate": 0.00028981790755306535, |
| "loss": 3.3137, |
| "step": 48100 |
| }, |
| { |
| "epoch": 5.182434614142719, |
| "grad_norm": 0.6586178541183472, |
| "learning_rate": 0.00028949466652300395, |
| "loss": 3.3183, |
| "step": 48150 |
| }, |
| { |
| "epoch": 5.187816166182327, |
| "grad_norm": 0.6953296661376953, |
| "learning_rate": 0.00028917142549294254, |
| "loss": 3.3212, |
| "step": 48200 |
| }, |
| { |
| "epoch": 5.193197718221935, |
| "grad_norm": 0.6340330839157104, |
| "learning_rate": 0.00028884818446288114, |
| "loss": 3.3081, |
| "step": 48250 |
| }, |
| { |
| "epoch": 5.198579270261543, |
| "grad_norm": 0.6623373627662659, |
| "learning_rate": 0.0002885249434328197, |
| "loss": 3.3145, |
| "step": 48300 |
| }, |
| { |
| "epoch": 5.203960822301152, |
| "grad_norm": 0.6238620281219482, |
| "learning_rate": 0.0002882017024027583, |
| "loss": 3.3138, |
| "step": 48350 |
| }, |
| { |
| "epoch": 5.20934237434076, |
| "grad_norm": 0.6164125204086304, |
| "learning_rate": 0.00028787846137269687, |
| "loss": 3.3321, |
| "step": 48400 |
| }, |
| { |
| "epoch": 5.214723926380368, |
| "grad_norm": 0.6289418339729309, |
| "learning_rate": 0.00028755522034263546, |
| "loss": 3.3213, |
| "step": 48450 |
| }, |
| { |
| "epoch": 5.220105478419977, |
| "grad_norm": 0.7005164623260498, |
| "learning_rate": 0.00028723197931257406, |
| "loss": 3.3121, |
| "step": 48500 |
| }, |
| { |
| "epoch": 5.225487030459584, |
| "grad_norm": 0.6364693641662598, |
| "learning_rate": 0.00028690873828251265, |
| "loss": 3.3398, |
| "step": 48550 |
| }, |
| { |
| "epoch": 5.230868582499193, |
| "grad_norm": 0.6689774394035339, |
| "learning_rate": 0.00028658549725245125, |
| "loss": 3.305, |
| "step": 48600 |
| }, |
| { |
| "epoch": 5.236250134538801, |
| "grad_norm": 0.6482794880867004, |
| "learning_rate": 0.0002862622562223898, |
| "loss": 3.3205, |
| "step": 48650 |
| }, |
| { |
| "epoch": 5.241631686578409, |
| "grad_norm": 0.6583704352378845, |
| "learning_rate": 0.0002859390151923284, |
| "loss": 3.3276, |
| "step": 48700 |
| }, |
| { |
| "epoch": 5.247013238618018, |
| "grad_norm": 0.7061691880226135, |
| "learning_rate": 0.000285615774162267, |
| "loss": 3.3136, |
| "step": 48750 |
| }, |
| { |
| "epoch": 5.252394790657625, |
| "grad_norm": 0.660162091255188, |
| "learning_rate": 0.00028529253313220557, |
| "loss": 3.3249, |
| "step": 48800 |
| }, |
| { |
| "epoch": 5.257776342697234, |
| "grad_norm": 0.7076647877693176, |
| "learning_rate": 0.0002849692921021441, |
| "loss": 3.3087, |
| "step": 48850 |
| }, |
| { |
| "epoch": 5.2631578947368425, |
| "grad_norm": 0.6797081828117371, |
| "learning_rate": 0.00028464605107208276, |
| "loss": 3.335, |
| "step": 48900 |
| }, |
| { |
| "epoch": 5.26853944677645, |
| "grad_norm": 0.6778848171234131, |
| "learning_rate": 0.0002843228100420213, |
| "loss": 3.316, |
| "step": 48950 |
| }, |
| { |
| "epoch": 5.273920998816059, |
| "grad_norm": 0.6917376518249512, |
| "learning_rate": 0.0002839995690119599, |
| "loss": 3.3007, |
| "step": 49000 |
| }, |
| { |
| "epoch": 5.273920998816059, |
| "eval_accuracy": 0.38130709641346594, |
| "eval_loss": 3.407010793685913, |
| "eval_runtime": 180.5321, |
| "eval_samples_per_second": 99.766, |
| "eval_steps_per_second": 6.237, |
| "step": 49000 |
| }, |
| { |
| "epoch": 5.279302550855666, |
| "grad_norm": 0.6813348531723022, |
| "learning_rate": 0.0002836763279818985, |
| "loss": 3.3281, |
| "step": 49050 |
| }, |
| { |
| "epoch": 5.284684102895275, |
| "grad_norm": 0.6178494095802307, |
| "learning_rate": 0.0002833530869518371, |
| "loss": 3.3281, |
| "step": 49100 |
| }, |
| { |
| "epoch": 5.2900656549348835, |
| "grad_norm": 0.7000295519828796, |
| "learning_rate": 0.0002830298459217756, |
| "loss": 3.3227, |
| "step": 49150 |
| }, |
| { |
| "epoch": 5.295447206974491, |
| "grad_norm": 0.6339635252952576, |
| "learning_rate": 0.0002827066048917142, |
| "loss": 3.3207, |
| "step": 49200 |
| }, |
| { |
| "epoch": 5.3008287590141, |
| "grad_norm": 0.6519315838813782, |
| "learning_rate": 0.0002823833638616528, |
| "loss": 3.3175, |
| "step": 49250 |
| }, |
| { |
| "epoch": 5.306210311053708, |
| "grad_norm": 0.7280599474906921, |
| "learning_rate": 0.0002820601228315914, |
| "loss": 3.3503, |
| "step": 49300 |
| }, |
| { |
| "epoch": 5.311591863093316, |
| "grad_norm": 0.6790415048599243, |
| "learning_rate": 0.00028173688180153, |
| "loss": 3.3403, |
| "step": 49350 |
| }, |
| { |
| "epoch": 5.316973415132924, |
| "grad_norm": 0.6788349151611328, |
| "learning_rate": 0.00028141364077146854, |
| "loss": 3.3234, |
| "step": 49400 |
| }, |
| { |
| "epoch": 5.322354967172533, |
| "grad_norm": 0.702439546585083, |
| "learning_rate": 0.00028109039974140714, |
| "loss": 3.335, |
| "step": 49450 |
| }, |
| { |
| "epoch": 5.327736519212141, |
| "grad_norm": 0.6609926223754883, |
| "learning_rate": 0.00028076715871134573, |
| "loss": 3.3331, |
| "step": 49500 |
| }, |
| { |
| "epoch": 5.333118071251749, |
| "grad_norm": 0.6514111757278442, |
| "learning_rate": 0.00028044391768128433, |
| "loss": 3.3408, |
| "step": 49550 |
| }, |
| { |
| "epoch": 5.338499623291357, |
| "grad_norm": 0.6543302536010742, |
| "learning_rate": 0.0002801206766512229, |
| "loss": 3.3421, |
| "step": 49600 |
| }, |
| { |
| "epoch": 5.343881175330965, |
| "grad_norm": 0.6633315086364746, |
| "learning_rate": 0.0002797974356211615, |
| "loss": 3.3449, |
| "step": 49650 |
| }, |
| { |
| "epoch": 5.349262727370574, |
| "grad_norm": 0.6488233804702759, |
| "learning_rate": 0.00027947419459110006, |
| "loss": 3.3186, |
| "step": 49700 |
| }, |
| { |
| "epoch": 5.354644279410182, |
| "grad_norm": 0.7404241561889648, |
| "learning_rate": 0.00027915095356103865, |
| "loss": 3.3246, |
| "step": 49750 |
| }, |
| { |
| "epoch": 5.36002583144979, |
| "grad_norm": 0.6592351198196411, |
| "learning_rate": 0.00027882771253097725, |
| "loss": 3.3266, |
| "step": 49800 |
| }, |
| { |
| "epoch": 5.365407383489399, |
| "grad_norm": 0.6456713676452637, |
| "learning_rate": 0.00027850447150091584, |
| "loss": 3.3199, |
| "step": 49850 |
| }, |
| { |
| "epoch": 5.370788935529006, |
| "grad_norm": 0.6827576160430908, |
| "learning_rate": 0.00027818769529145565, |
| "loss": 3.3212, |
| "step": 49900 |
| }, |
| { |
| "epoch": 5.376170487568615, |
| "grad_norm": 0.6642194390296936, |
| "learning_rate": 0.0002778644542613942, |
| "loss": 3.3367, |
| "step": 49950 |
| }, |
| { |
| "epoch": 5.3815520396082235, |
| "grad_norm": 0.6693667769432068, |
| "learning_rate": 0.00027754121323133284, |
| "loss": 3.3337, |
| "step": 50000 |
| }, |
| { |
| "epoch": 5.3815520396082235, |
| "eval_accuracy": 0.38180950736672015, |
| "eval_loss": 3.403062343597412, |
| "eval_runtime": 180.255, |
| "eval_samples_per_second": 99.92, |
| "eval_steps_per_second": 6.247, |
| "step": 50000 |
| }, |
| { |
| "epoch": 5.386933591647831, |
| "grad_norm": 0.6618473529815674, |
| "learning_rate": 0.0002772179722012714, |
| "loss": 3.3359, |
| "step": 50050 |
| }, |
| { |
| "epoch": 5.39231514368744, |
| "grad_norm": 0.6960184574127197, |
| "learning_rate": 0.00027689473117120997, |
| "loss": 3.3385, |
| "step": 50100 |
| }, |
| { |
| "epoch": 5.397696695727047, |
| "grad_norm": 0.6597319841384888, |
| "learning_rate": 0.00027657149014114857, |
| "loss": 3.3239, |
| "step": 50150 |
| }, |
| { |
| "epoch": 5.403078247766656, |
| "grad_norm": 0.6477595567703247, |
| "learning_rate": 0.00027624824911108716, |
| "loss": 3.337, |
| "step": 50200 |
| }, |
| { |
| "epoch": 5.4084597998062645, |
| "grad_norm": 0.6981415152549744, |
| "learning_rate": 0.0002759250080810257, |
| "loss": 3.3085, |
| "step": 50250 |
| }, |
| { |
| "epoch": 5.413841351845872, |
| "grad_norm": 0.6679623126983643, |
| "learning_rate": 0.0002756017670509643, |
| "loss": 3.3104, |
| "step": 50300 |
| }, |
| { |
| "epoch": 5.419222903885481, |
| "grad_norm": 0.6863991618156433, |
| "learning_rate": 0.0002752785260209029, |
| "loss": 3.3437, |
| "step": 50350 |
| }, |
| { |
| "epoch": 5.424604455925088, |
| "grad_norm": 0.6082434058189392, |
| "learning_rate": 0.0002749552849908415, |
| "loss": 3.3244, |
| "step": 50400 |
| }, |
| { |
| "epoch": 5.429986007964697, |
| "grad_norm": 0.6896102428436279, |
| "learning_rate": 0.0002746320439607801, |
| "loss": 3.3639, |
| "step": 50450 |
| }, |
| { |
| "epoch": 5.435367560004305, |
| "grad_norm": 0.688849151134491, |
| "learning_rate": 0.0002743088029307186, |
| "loss": 3.3353, |
| "step": 50500 |
| }, |
| { |
| "epoch": 5.440749112043913, |
| "grad_norm": 0.7596466541290283, |
| "learning_rate": 0.00027398556190065727, |
| "loss": 3.3247, |
| "step": 50550 |
| }, |
| { |
| "epoch": 5.446130664083522, |
| "grad_norm": 0.6528745293617249, |
| "learning_rate": 0.0002736623208705958, |
| "loss": 3.3308, |
| "step": 50600 |
| }, |
| { |
| "epoch": 5.45151221612313, |
| "grad_norm": 0.6976600885391235, |
| "learning_rate": 0.0002733390798405344, |
| "loss": 3.3428, |
| "step": 50650 |
| }, |
| { |
| "epoch": 5.456893768162738, |
| "grad_norm": 0.6412066221237183, |
| "learning_rate": 0.000273015838810473, |
| "loss": 3.337, |
| "step": 50700 |
| }, |
| { |
| "epoch": 5.462275320202346, |
| "grad_norm": 0.6509308815002441, |
| "learning_rate": 0.0002726925977804116, |
| "loss": 3.3395, |
| "step": 50750 |
| }, |
| { |
| "epoch": 5.467656872241955, |
| "grad_norm": 0.661870002746582, |
| "learning_rate": 0.00027236935675035013, |
| "loss": 3.3315, |
| "step": 50800 |
| }, |
| { |
| "epoch": 5.473038424281563, |
| "grad_norm": 0.6559614539146423, |
| "learning_rate": 0.00027204611572028873, |
| "loss": 3.3327, |
| "step": 50850 |
| }, |
| { |
| "epoch": 5.478419976321171, |
| "grad_norm": 0.7063403129577637, |
| "learning_rate": 0.0002717228746902273, |
| "loss": 3.3128, |
| "step": 50900 |
| }, |
| { |
| "epoch": 5.483801528360779, |
| "grad_norm": 0.6607415676116943, |
| "learning_rate": 0.0002713996336601659, |
| "loss": 3.3446, |
| "step": 50950 |
| }, |
| { |
| "epoch": 5.489183080400387, |
| "grad_norm": 0.6556530594825745, |
| "learning_rate": 0.0002710763926301045, |
| "loss": 3.3386, |
| "step": 51000 |
| }, |
| { |
| "epoch": 5.489183080400387, |
| "eval_accuracy": 0.3822826906931522, |
| "eval_loss": 3.398730516433716, |
| "eval_runtime": 180.1394, |
| "eval_samples_per_second": 99.984, |
| "eval_steps_per_second": 6.251, |
| "step": 51000 |
| }, |
| { |
| "epoch": 5.494564632439996, |
| "grad_norm": 0.6305342316627502, |
| "learning_rate": 0.00027075315160004305, |
| "loss": 3.3359, |
| "step": 51050 |
| }, |
| { |
| "epoch": 5.499946184479604, |
| "grad_norm": 0.7005680799484253, |
| "learning_rate": 0.00027042991056998165, |
| "loss": 3.3388, |
| "step": 51100 |
| }, |
| { |
| "epoch": 5.505327736519212, |
| "grad_norm": 0.6419593095779419, |
| "learning_rate": 0.00027010666953992024, |
| "loss": 3.3246, |
| "step": 51150 |
| }, |
| { |
| "epoch": 5.510709288558821, |
| "grad_norm": 0.6594645380973816, |
| "learning_rate": 0.00026978342850985884, |
| "loss": 3.3368, |
| "step": 51200 |
| }, |
| { |
| "epoch": 5.516090840598428, |
| "grad_norm": 0.6918158531188965, |
| "learning_rate": 0.0002694601874797974, |
| "loss": 3.3387, |
| "step": 51250 |
| }, |
| { |
| "epoch": 5.521472392638037, |
| "grad_norm": 0.6681293249130249, |
| "learning_rate": 0.000269136946449736, |
| "loss": 3.3247, |
| "step": 51300 |
| }, |
| { |
| "epoch": 5.5268539446776455, |
| "grad_norm": 0.6619294881820679, |
| "learning_rate": 0.00026881370541967457, |
| "loss": 3.3303, |
| "step": 51350 |
| }, |
| { |
| "epoch": 5.532235496717253, |
| "grad_norm": 0.6340798735618591, |
| "learning_rate": 0.00026849046438961316, |
| "loss": 3.3283, |
| "step": 51400 |
| }, |
| { |
| "epoch": 5.537617048756862, |
| "grad_norm": 0.715398371219635, |
| "learning_rate": 0.00026816722335955176, |
| "loss": 3.3455, |
| "step": 51450 |
| }, |
| { |
| "epoch": 5.542998600796469, |
| "grad_norm": 0.6638710498809814, |
| "learning_rate": 0.00026784398232949035, |
| "loss": 3.3138, |
| "step": 51500 |
| }, |
| { |
| "epoch": 5.548380152836078, |
| "grad_norm": 0.6782832741737366, |
| "learning_rate": 0.00026752074129942895, |
| "loss": 3.3473, |
| "step": 51550 |
| }, |
| { |
| "epoch": 5.553761704875686, |
| "grad_norm": 0.6959323287010193, |
| "learning_rate": 0.0002671975002693675, |
| "loss": 3.3321, |
| "step": 51600 |
| }, |
| { |
| "epoch": 5.559143256915294, |
| "grad_norm": 0.6669702529907227, |
| "learning_rate": 0.0002668742592393061, |
| "loss": 3.3203, |
| "step": 51650 |
| }, |
| { |
| "epoch": 5.564524808954903, |
| "grad_norm": 0.6613565683364868, |
| "learning_rate": 0.0002665510182092447, |
| "loss": 3.3511, |
| "step": 51700 |
| }, |
| { |
| "epoch": 5.569906360994511, |
| "grad_norm": 0.6775053143501282, |
| "learning_rate": 0.00026622777717918327, |
| "loss": 3.3312, |
| "step": 51750 |
| }, |
| { |
| "epoch": 5.575287913034119, |
| "grad_norm": 0.6518484354019165, |
| "learning_rate": 0.0002659045361491218, |
| "loss": 3.3478, |
| "step": 51800 |
| }, |
| { |
| "epoch": 5.580669465073727, |
| "grad_norm": 0.718291699886322, |
| "learning_rate": 0.00026558129511906046, |
| "loss": 3.3449, |
| "step": 51850 |
| }, |
| { |
| "epoch": 5.586051017113336, |
| "grad_norm": 0.6551558971405029, |
| "learning_rate": 0.000265258054088999, |
| "loss": 3.3506, |
| "step": 51900 |
| }, |
| { |
| "epoch": 5.591432569152944, |
| "grad_norm": 0.622401773929596, |
| "learning_rate": 0.0002649412778795388, |
| "loss": 3.3253, |
| "step": 51950 |
| }, |
| { |
| "epoch": 5.596814121192552, |
| "grad_norm": 0.6426682472229004, |
| "learning_rate": 0.0002646180368494774, |
| "loss": 3.3294, |
| "step": 52000 |
| }, |
| { |
| "epoch": 5.596814121192552, |
| "eval_accuracy": 0.3830105563886214, |
| "eval_loss": 3.3924288749694824, |
| "eval_runtime": 180.6606, |
| "eval_samples_per_second": 99.695, |
| "eval_steps_per_second": 6.233, |
| "step": 52000 |
| }, |
| { |
| "epoch": 5.60219567323216, |
| "grad_norm": 0.6933045387268066, |
| "learning_rate": 0.000264294795819416, |
| "loss": 3.3382, |
| "step": 52050 |
| }, |
| { |
| "epoch": 5.607577225271768, |
| "grad_norm": 0.6460835933685303, |
| "learning_rate": 0.0002639715547893546, |
| "loss": 3.3259, |
| "step": 52100 |
| }, |
| { |
| "epoch": 5.612958777311377, |
| "grad_norm": 0.718722403049469, |
| "learning_rate": 0.00026364831375929313, |
| "loss": 3.342, |
| "step": 52150 |
| }, |
| { |
| "epoch": 5.618340329350985, |
| "grad_norm": 0.6578804850578308, |
| "learning_rate": 0.0002633250727292318, |
| "loss": 3.3247, |
| "step": 52200 |
| }, |
| { |
| "epoch": 5.623721881390593, |
| "grad_norm": 0.7196077704429626, |
| "learning_rate": 0.0002630018316991703, |
| "loss": 3.3305, |
| "step": 52250 |
| }, |
| { |
| "epoch": 5.629103433430201, |
| "grad_norm": 0.7222257256507874, |
| "learning_rate": 0.0002626785906691089, |
| "loss": 3.3448, |
| "step": 52300 |
| }, |
| { |
| "epoch": 5.634484985469809, |
| "grad_norm": 0.6821712255477905, |
| "learning_rate": 0.0002623553496390475, |
| "loss": 3.3404, |
| "step": 52350 |
| }, |
| { |
| "epoch": 5.639866537509418, |
| "grad_norm": 0.6659912467002869, |
| "learning_rate": 0.0002620321086089861, |
| "loss": 3.3263, |
| "step": 52400 |
| }, |
| { |
| "epoch": 5.645248089549026, |
| "grad_norm": 0.6843506097793579, |
| "learning_rate": 0.00026170886757892464, |
| "loss": 3.316, |
| "step": 52450 |
| }, |
| { |
| "epoch": 5.650629641588634, |
| "grad_norm": 0.6823989748954773, |
| "learning_rate": 0.00026138562654886324, |
| "loss": 3.3422, |
| "step": 52500 |
| }, |
| { |
| "epoch": 5.656011193628243, |
| "grad_norm": 0.6962025761604309, |
| "learning_rate": 0.00026106238551880183, |
| "loss": 3.3361, |
| "step": 52550 |
| }, |
| { |
| "epoch": 5.66139274566785, |
| "grad_norm": 0.6766282916069031, |
| "learning_rate": 0.00026073914448874043, |
| "loss": 3.328, |
| "step": 52600 |
| }, |
| { |
| "epoch": 5.666774297707459, |
| "grad_norm": 0.6884021759033203, |
| "learning_rate": 0.000260415903458679, |
| "loss": 3.3356, |
| "step": 52650 |
| }, |
| { |
| "epoch": 5.672155849747067, |
| "grad_norm": 0.6552942395210266, |
| "learning_rate": 0.00026009266242861756, |
| "loss": 3.3326, |
| "step": 52700 |
| }, |
| { |
| "epoch": 5.677537401786675, |
| "grad_norm": 0.649552583694458, |
| "learning_rate": 0.00025976942139855616, |
| "loss": 3.3303, |
| "step": 52750 |
| }, |
| { |
| "epoch": 5.682918953826284, |
| "grad_norm": 0.7173223495483398, |
| "learning_rate": 0.00025945264518909596, |
| "loss": 3.3183, |
| "step": 52800 |
| }, |
| { |
| "epoch": 5.688300505865891, |
| "grad_norm": 0.7656635046005249, |
| "learning_rate": 0.00025912940415903456, |
| "loss": 3.3251, |
| "step": 52850 |
| }, |
| { |
| "epoch": 5.6936820579055, |
| "grad_norm": 0.740873396396637, |
| "learning_rate": 0.00025880616312897315, |
| "loss": 3.3179, |
| "step": 52900 |
| }, |
| { |
| "epoch": 5.699063609945108, |
| "grad_norm": 0.7117483019828796, |
| "learning_rate": 0.00025848292209891175, |
| "loss": 3.3273, |
| "step": 52950 |
| }, |
| { |
| "epoch": 5.704445161984716, |
| "grad_norm": 0.68827223777771, |
| "learning_rate": 0.00025815968106885034, |
| "loss": 3.3153, |
| "step": 53000 |
| }, |
| { |
| "epoch": 5.704445161984716, |
| "eval_accuracy": 0.38329913845865665, |
| "eval_loss": 3.3908591270446777, |
| "eval_runtime": 180.1829, |
| "eval_samples_per_second": 99.96, |
| "eval_steps_per_second": 6.249, |
| "step": 53000 |
| }, |
| { |
| "epoch": 5.709826714024325, |
| "grad_norm": 0.6453783512115479, |
| "learning_rate": 0.0002578364400387889, |
| "loss": 3.3325, |
| "step": 53050 |
| }, |
| { |
| "epoch": 5.715208266063933, |
| "grad_norm": 0.696941077709198, |
| "learning_rate": 0.0002575131990087275, |
| "loss": 3.3351, |
| "step": 53100 |
| }, |
| { |
| "epoch": 5.720589818103541, |
| "grad_norm": 0.6282665133476257, |
| "learning_rate": 0.0002571899579786661, |
| "loss": 3.3288, |
| "step": 53150 |
| }, |
| { |
| "epoch": 5.725971370143149, |
| "grad_norm": 0.6461380124092102, |
| "learning_rate": 0.00025686671694860467, |
| "loss": 3.3403, |
| "step": 53200 |
| }, |
| { |
| "epoch": 5.731352922182758, |
| "grad_norm": 0.6844844818115234, |
| "learning_rate": 0.0002565434759185432, |
| "loss": 3.3306, |
| "step": 53250 |
| }, |
| { |
| "epoch": 5.736734474222366, |
| "grad_norm": 0.6883606314659119, |
| "learning_rate": 0.00025622023488848186, |
| "loss": 3.3242, |
| "step": 53300 |
| }, |
| { |
| "epoch": 5.742116026261974, |
| "grad_norm": 0.7068212628364563, |
| "learning_rate": 0.0002558969938584204, |
| "loss": 3.3241, |
| "step": 53350 |
| }, |
| { |
| "epoch": 5.747497578301582, |
| "grad_norm": 0.7148516774177551, |
| "learning_rate": 0.000255573752828359, |
| "loss": 3.326, |
| "step": 53400 |
| }, |
| { |
| "epoch": 5.75287913034119, |
| "grad_norm": 0.6874505877494812, |
| "learning_rate": 0.0002552505117982976, |
| "loss": 3.3322, |
| "step": 53450 |
| }, |
| { |
| "epoch": 5.758260682380799, |
| "grad_norm": 0.655957818031311, |
| "learning_rate": 0.0002549272707682362, |
| "loss": 3.3385, |
| "step": 53500 |
| }, |
| { |
| "epoch": 5.763642234420407, |
| "grad_norm": 0.6629857420921326, |
| "learning_rate": 0.0002546040297381747, |
| "loss": 3.3132, |
| "step": 53550 |
| }, |
| { |
| "epoch": 5.769023786460015, |
| "grad_norm": 0.671413779258728, |
| "learning_rate": 0.0002542807887081133, |
| "loss": 3.3402, |
| "step": 53600 |
| }, |
| { |
| "epoch": 5.774405338499624, |
| "grad_norm": 0.6932768225669861, |
| "learning_rate": 0.0002539575476780519, |
| "loss": 3.3157, |
| "step": 53650 |
| }, |
| { |
| "epoch": 5.779786890539231, |
| "grad_norm": 0.6655086874961853, |
| "learning_rate": 0.0002536343066479905, |
| "loss": 3.3101, |
| "step": 53700 |
| }, |
| { |
| "epoch": 5.78516844257884, |
| "grad_norm": 0.6800814270973206, |
| "learning_rate": 0.0002533110656179291, |
| "loss": 3.3382, |
| "step": 53750 |
| }, |
| { |
| "epoch": 5.790549994618448, |
| "grad_norm": 0.665797233581543, |
| "learning_rate": 0.00025298782458786764, |
| "loss": 3.3159, |
| "step": 53800 |
| }, |
| { |
| "epoch": 5.795931546658056, |
| "grad_norm": 0.701047956943512, |
| "learning_rate": 0.00025266458355780624, |
| "loss": 3.3319, |
| "step": 53850 |
| }, |
| { |
| "epoch": 5.801313098697665, |
| "grad_norm": 0.6783666014671326, |
| "learning_rate": 0.00025234134252774483, |
| "loss": 3.3272, |
| "step": 53900 |
| }, |
| { |
| "epoch": 5.806694650737272, |
| "grad_norm": 0.6685265302658081, |
| "learning_rate": 0.0002520181014976834, |
| "loss": 3.3389, |
| "step": 53950 |
| }, |
| { |
| "epoch": 5.812076202776881, |
| "grad_norm": 0.6971936821937561, |
| "learning_rate": 0.000251694860467622, |
| "loss": 3.3324, |
| "step": 54000 |
| }, |
| { |
| "epoch": 5.812076202776881, |
| "eval_accuracy": 0.3839824564701219, |
| "eval_loss": 3.3846652507781982, |
| "eval_runtime": 180.3207, |
| "eval_samples_per_second": 99.883, |
| "eval_steps_per_second": 6.244, |
| "step": 54000 |
| }, |
| { |
| "epoch": 5.817457754816489, |
| "grad_norm": 0.6983776092529297, |
| "learning_rate": 0.0002513716194375606, |
| "loss": 3.3241, |
| "step": 54050 |
| }, |
| { |
| "epoch": 5.822839306856097, |
| "grad_norm": 0.6729536056518555, |
| "learning_rate": 0.00025104837840749916, |
| "loss": 3.3321, |
| "step": 54100 |
| }, |
| { |
| "epoch": 5.828220858895706, |
| "grad_norm": 0.6928292512893677, |
| "learning_rate": 0.00025072513737743775, |
| "loss": 3.3442, |
| "step": 54150 |
| }, |
| { |
| "epoch": 5.833602410935313, |
| "grad_norm": 0.6832127571105957, |
| "learning_rate": 0.00025040189634737634, |
| "loss": 3.33, |
| "step": 54200 |
| }, |
| { |
| "epoch": 5.838983962974922, |
| "grad_norm": 0.7145037651062012, |
| "learning_rate": 0.00025007865531731494, |
| "loss": 3.3557, |
| "step": 54250 |
| }, |
| { |
| "epoch": 5.84436551501453, |
| "grad_norm": 0.7388347387313843, |
| "learning_rate": 0.00024975541428725353, |
| "loss": 3.3417, |
| "step": 54300 |
| }, |
| { |
| "epoch": 5.849747067054138, |
| "grad_norm": 0.6939401030540466, |
| "learning_rate": 0.0002494321732571921, |
| "loss": 3.3285, |
| "step": 54350 |
| }, |
| { |
| "epoch": 5.855128619093747, |
| "grad_norm": 0.7111252546310425, |
| "learning_rate": 0.00024910893222713067, |
| "loss": 3.3386, |
| "step": 54400 |
| }, |
| { |
| "epoch": 5.860510171133355, |
| "grad_norm": 0.7177287936210632, |
| "learning_rate": 0.00024878569119706926, |
| "loss": 3.3222, |
| "step": 54450 |
| }, |
| { |
| "epoch": 5.865891723172963, |
| "grad_norm": 0.6975011229515076, |
| "learning_rate": 0.00024846245016700786, |
| "loss": 3.3301, |
| "step": 54500 |
| }, |
| { |
| "epoch": 5.871273275212571, |
| "grad_norm": 0.8165584802627563, |
| "learning_rate": 0.0002481392091369464, |
| "loss": 3.3396, |
| "step": 54550 |
| }, |
| { |
| "epoch": 5.87665482725218, |
| "grad_norm": 0.7520706653594971, |
| "learning_rate": 0.00024781596810688505, |
| "loss": 3.3314, |
| "step": 54600 |
| }, |
| { |
| "epoch": 5.882036379291788, |
| "grad_norm": 0.6771239042282104, |
| "learning_rate": 0.0002474927270768236, |
| "loss": 3.3574, |
| "step": 54650 |
| }, |
| { |
| "epoch": 5.887417931331396, |
| "grad_norm": 0.6594865918159485, |
| "learning_rate": 0.0002471694860467622, |
| "loss": 3.3361, |
| "step": 54700 |
| }, |
| { |
| "epoch": 5.892799483371004, |
| "grad_norm": 0.6948037147521973, |
| "learning_rate": 0.0002468462450167008, |
| "loss": 3.3186, |
| "step": 54750 |
| }, |
| { |
| "epoch": 5.898181035410612, |
| "grad_norm": 0.680067241191864, |
| "learning_rate": 0.0002465230039866393, |
| "loss": 3.3348, |
| "step": 54800 |
| }, |
| { |
| "epoch": 5.903562587450221, |
| "grad_norm": 0.6603071689605713, |
| "learning_rate": 0.0002461997629565779, |
| "loss": 3.3266, |
| "step": 54850 |
| }, |
| { |
| "epoch": 5.9089441394898286, |
| "grad_norm": 0.6521521210670471, |
| "learning_rate": 0.0002458765219265165, |
| "loss": 3.3288, |
| "step": 54900 |
| }, |
| { |
| "epoch": 5.914325691529437, |
| "grad_norm": 0.8012475371360779, |
| "learning_rate": 0.0002455532808964551, |
| "loss": 3.3194, |
| "step": 54950 |
| }, |
| { |
| "epoch": 5.919707243569046, |
| "grad_norm": 0.7362468242645264, |
| "learning_rate": 0.0002452300398663937, |
| "loss": 3.3298, |
| "step": 55000 |
| }, |
| { |
| "epoch": 5.919707243569046, |
| "eval_accuracy": 0.3843108054970973, |
| "eval_loss": 3.38219952583313, |
| "eval_runtime": 180.5029, |
| "eval_samples_per_second": 99.782, |
| "eval_steps_per_second": 6.238, |
| "step": 55000 |
| }, |
| { |
| "epoch": 5.925088795608653, |
| "grad_norm": 0.7196722030639648, |
| "learning_rate": 0.0002449067988363323, |
| "loss": 3.3229, |
| "step": 55050 |
| }, |
| { |
| "epoch": 5.930470347648262, |
| "grad_norm": 0.7012231945991516, |
| "learning_rate": 0.00024458355780627083, |
| "loss": 3.3314, |
| "step": 55100 |
| }, |
| { |
| "epoch": 5.93585189968787, |
| "grad_norm": 0.7560854554176331, |
| "learning_rate": 0.0002442603167762094, |
| "loss": 3.3366, |
| "step": 55150 |
| }, |
| { |
| "epoch": 5.941233451727478, |
| "grad_norm": 0.6536340713500977, |
| "learning_rate": 0.00024393707574614802, |
| "loss": 3.3336, |
| "step": 55200 |
| }, |
| { |
| "epoch": 5.946615003767087, |
| "grad_norm": 0.7389876246452332, |
| "learning_rate": 0.0002436138347160866, |
| "loss": 3.3283, |
| "step": 55250 |
| }, |
| { |
| "epoch": 5.951996555806694, |
| "grad_norm": 0.6520614624023438, |
| "learning_rate": 0.0002432905936860252, |
| "loss": 3.3385, |
| "step": 55300 |
| }, |
| { |
| "epoch": 5.957378107846303, |
| "grad_norm": 0.6752536296844482, |
| "learning_rate": 0.00024296735265596378, |
| "loss": 3.3286, |
| "step": 55350 |
| }, |
| { |
| "epoch": 5.962759659885911, |
| "grad_norm": 0.6813256740570068, |
| "learning_rate": 0.00024264411162590235, |
| "loss": 3.3388, |
| "step": 55400 |
| }, |
| { |
| "epoch": 5.968141211925519, |
| "grad_norm": 0.6881508827209473, |
| "learning_rate": 0.00024232087059584097, |
| "loss": 3.325, |
| "step": 55450 |
| }, |
| { |
| "epoch": 5.973522763965128, |
| "grad_norm": 0.7486839890480042, |
| "learning_rate": 0.00024199762956577953, |
| "loss": 3.3334, |
| "step": 55500 |
| }, |
| { |
| "epoch": 5.978904316004736, |
| "grad_norm": 0.6768556237220764, |
| "learning_rate": 0.0002416743885357181, |
| "loss": 3.331, |
| "step": 55550 |
| }, |
| { |
| "epoch": 5.984285868044344, |
| "grad_norm": 0.7191193699836731, |
| "learning_rate": 0.0002413511475056567, |
| "loss": 3.3214, |
| "step": 55600 |
| }, |
| { |
| "epoch": 5.989667420083952, |
| "grad_norm": 0.7417619228363037, |
| "learning_rate": 0.0002410279064755953, |
| "loss": 3.3383, |
| "step": 55650 |
| }, |
| { |
| "epoch": 5.995048972123561, |
| "grad_norm": 0.7486613392829895, |
| "learning_rate": 0.00024070466544553386, |
| "loss": 3.3266, |
| "step": 55700 |
| }, |
| { |
| "epoch": 6.000430524163169, |
| "grad_norm": 0.6875068545341492, |
| "learning_rate": 0.00024038142441547245, |
| "loss": 3.3292, |
| "step": 55750 |
| }, |
| { |
| "epoch": 6.005812076202777, |
| "grad_norm": 0.7089745998382568, |
| "learning_rate": 0.00024005818338541102, |
| "loss": 3.2554, |
| "step": 55800 |
| }, |
| { |
| "epoch": 6.011193628242385, |
| "grad_norm": 0.6928079724311829, |
| "learning_rate": 0.0002397349423553496, |
| "loss": 3.2292, |
| "step": 55850 |
| }, |
| { |
| "epoch": 6.016575180281993, |
| "grad_norm": 0.6740532517433167, |
| "learning_rate": 0.0002394117013252882, |
| "loss": 3.2425, |
| "step": 55900 |
| }, |
| { |
| "epoch": 6.021956732321602, |
| "grad_norm": 0.6799925565719604, |
| "learning_rate": 0.00023908846029522678, |
| "loss": 3.2304, |
| "step": 55950 |
| }, |
| { |
| "epoch": 6.0273382843612096, |
| "grad_norm": 0.7598007917404175, |
| "learning_rate": 0.00023876521926516537, |
| "loss": 3.2505, |
| "step": 56000 |
| }, |
| { |
| "epoch": 6.0273382843612096, |
| "eval_accuracy": 0.3845337612228926, |
| "eval_loss": 3.3847992420196533, |
| "eval_runtime": 180.0957, |
| "eval_samples_per_second": 100.008, |
| "eval_steps_per_second": 6.252, |
| "step": 56000 |
| }, |
| { |
| "epoch": 6.032719836400818, |
| "grad_norm": 0.7275548577308655, |
| "learning_rate": 0.00023844197823510397, |
| "loss": 3.248, |
| "step": 56050 |
| }, |
| { |
| "epoch": 6.038101388440427, |
| "grad_norm": 0.7083311676979065, |
| "learning_rate": 0.00023811873720504253, |
| "loss": 3.2203, |
| "step": 56100 |
| }, |
| { |
| "epoch": 6.043482940480034, |
| "grad_norm": 0.7197110056877136, |
| "learning_rate": 0.00023779549617498113, |
| "loss": 3.2358, |
| "step": 56150 |
| }, |
| { |
| "epoch": 6.048864492519643, |
| "grad_norm": 0.6651777029037476, |
| "learning_rate": 0.0002374722551449197, |
| "loss": 3.2521, |
| "step": 56200 |
| }, |
| { |
| "epoch": 6.0542460445592505, |
| "grad_norm": 0.7160053253173828, |
| "learning_rate": 0.0002371490141148583, |
| "loss": 3.2404, |
| "step": 56250 |
| }, |
| { |
| "epoch": 6.059627596598859, |
| "grad_norm": 0.7208572626113892, |
| "learning_rate": 0.00023682577308479689, |
| "loss": 3.244, |
| "step": 56300 |
| }, |
| { |
| "epoch": 6.065009148638468, |
| "grad_norm": 0.6583945155143738, |
| "learning_rate": 0.00023650253205473545, |
| "loss": 3.2551, |
| "step": 56350 |
| }, |
| { |
| "epoch": 6.070390700678075, |
| "grad_norm": 0.695590078830719, |
| "learning_rate": 0.00023617929102467402, |
| "loss": 3.25, |
| "step": 56400 |
| }, |
| { |
| "epoch": 6.075772252717684, |
| "grad_norm": 0.6976557374000549, |
| "learning_rate": 0.00023586251481521386, |
| "loss": 3.2608, |
| "step": 56450 |
| }, |
| { |
| "epoch": 6.081153804757292, |
| "grad_norm": 0.7817869782447815, |
| "learning_rate": 0.00023553927378515242, |
| "loss": 3.2698, |
| "step": 56500 |
| }, |
| { |
| "epoch": 6.0865353567969, |
| "grad_norm": 0.6705769300460815, |
| "learning_rate": 0.00023521603275509104, |
| "loss": 3.2602, |
| "step": 56550 |
| }, |
| { |
| "epoch": 6.091916908836509, |
| "grad_norm": 0.7016733884811401, |
| "learning_rate": 0.0002348927917250296, |
| "loss": 3.2514, |
| "step": 56600 |
| }, |
| { |
| "epoch": 6.097298460876116, |
| "grad_norm": 0.7753743529319763, |
| "learning_rate": 0.00023456955069496818, |
| "loss": 3.2367, |
| "step": 56650 |
| }, |
| { |
| "epoch": 6.102680012915725, |
| "grad_norm": 0.6739595532417297, |
| "learning_rate": 0.00023424630966490677, |
| "loss": 3.2541, |
| "step": 56700 |
| }, |
| { |
| "epoch": 6.108061564955333, |
| "grad_norm": 0.732443630695343, |
| "learning_rate": 0.00023392306863484537, |
| "loss": 3.2537, |
| "step": 56750 |
| }, |
| { |
| "epoch": 6.113443116994941, |
| "grad_norm": 0.7694481611251831, |
| "learning_rate": 0.00023359982760478396, |
| "loss": 3.2594, |
| "step": 56800 |
| }, |
| { |
| "epoch": 6.11882466903455, |
| "grad_norm": 0.7056264877319336, |
| "learning_rate": 0.00023327658657472253, |
| "loss": 3.2455, |
| "step": 56850 |
| }, |
| { |
| "epoch": 6.124206221074158, |
| "grad_norm": 0.6392900943756104, |
| "learning_rate": 0.0002329533455446611, |
| "loss": 3.2531, |
| "step": 56900 |
| }, |
| { |
| "epoch": 6.129587773113766, |
| "grad_norm": 0.6789774298667908, |
| "learning_rate": 0.00023263010451459972, |
| "loss": 3.2526, |
| "step": 56950 |
| }, |
| { |
| "epoch": 6.134969325153374, |
| "grad_norm": 0.7180214524269104, |
| "learning_rate": 0.0002323068634845383, |
| "loss": 3.2374, |
| "step": 57000 |
| }, |
| { |
| "epoch": 6.134969325153374, |
| "eval_accuracy": 0.3843350350910799, |
| "eval_loss": 3.387458562850952, |
| "eval_runtime": 179.945, |
| "eval_samples_per_second": 100.092, |
| "eval_steps_per_second": 6.257, |
| "step": 57000 |
| }, |
| { |
| "epoch": 6.140350877192983, |
| "grad_norm": 0.6634199619293213, |
| "learning_rate": 0.00023198362245447686, |
| "loss": 3.2663, |
| "step": 57050 |
| }, |
| { |
| "epoch": 6.1457324292325906, |
| "grad_norm": 0.7024934887886047, |
| "learning_rate": 0.00023166038142441548, |
| "loss": 3.2761, |
| "step": 57100 |
| }, |
| { |
| "epoch": 6.151113981272199, |
| "grad_norm": 0.7180144190788269, |
| "learning_rate": 0.00023133714039435404, |
| "loss": 3.2654, |
| "step": 57150 |
| }, |
| { |
| "epoch": 6.156495533311807, |
| "grad_norm": 0.6728076934814453, |
| "learning_rate": 0.0002310138993642926, |
| "loss": 3.2579, |
| "step": 57200 |
| }, |
| { |
| "epoch": 6.161877085351415, |
| "grad_norm": 0.7283917665481567, |
| "learning_rate": 0.0002306906583342312, |
| "loss": 3.2565, |
| "step": 57250 |
| }, |
| { |
| "epoch": 6.167258637391024, |
| "grad_norm": 0.7351763248443604, |
| "learning_rate": 0.00023036741730416977, |
| "loss": 3.2523, |
| "step": 57300 |
| }, |
| { |
| "epoch": 6.1726401894306315, |
| "grad_norm": 0.7366397380828857, |
| "learning_rate": 0.00023004417627410837, |
| "loss": 3.2665, |
| "step": 57350 |
| }, |
| { |
| "epoch": 6.17802174147024, |
| "grad_norm": 0.6817381381988525, |
| "learning_rate": 0.00022972093524404696, |
| "loss": 3.2791, |
| "step": 57400 |
| }, |
| { |
| "epoch": 6.183403293509849, |
| "grad_norm": 0.6933722496032715, |
| "learning_rate": 0.00022939769421398553, |
| "loss": 3.2675, |
| "step": 57450 |
| }, |
| { |
| "epoch": 6.188784845549456, |
| "grad_norm": 0.6675575375556946, |
| "learning_rate": 0.0002290744531839241, |
| "loss": 3.252, |
| "step": 57500 |
| }, |
| { |
| "epoch": 6.194166397589065, |
| "grad_norm": 0.7353582382202148, |
| "learning_rate": 0.00022875121215386272, |
| "loss": 3.2649, |
| "step": 57550 |
| }, |
| { |
| "epoch": 6.1995479496286725, |
| "grad_norm": 0.7077396512031555, |
| "learning_rate": 0.0002284279711238013, |
| "loss": 3.2762, |
| "step": 57600 |
| }, |
| { |
| "epoch": 6.204929501668281, |
| "grad_norm": 0.761920154094696, |
| "learning_rate": 0.00022810473009373986, |
| "loss": 3.2494, |
| "step": 57650 |
| }, |
| { |
| "epoch": 6.21031105370789, |
| "grad_norm": 0.6998116970062256, |
| "learning_rate": 0.00022778148906367848, |
| "loss": 3.2744, |
| "step": 57700 |
| }, |
| { |
| "epoch": 6.215692605747497, |
| "grad_norm": 0.6889002919197083, |
| "learning_rate": 0.00022745824803361705, |
| "loss": 3.2544, |
| "step": 57750 |
| }, |
| { |
| "epoch": 6.221074157787106, |
| "grad_norm": 0.7440342307090759, |
| "learning_rate": 0.00022713500700355564, |
| "loss": 3.2774, |
| "step": 57800 |
| }, |
| { |
| "epoch": 6.226455709826714, |
| "grad_norm": 0.6621478199958801, |
| "learning_rate": 0.0002268117659734942, |
| "loss": 3.2952, |
| "step": 57850 |
| }, |
| { |
| "epoch": 6.231837261866322, |
| "grad_norm": 0.7194118499755859, |
| "learning_rate": 0.0002264885249434328, |
| "loss": 3.2817, |
| "step": 57900 |
| }, |
| { |
| "epoch": 6.237218813905931, |
| "grad_norm": 0.7072327136993408, |
| "learning_rate": 0.0002261652839133714, |
| "loss": 3.2649, |
| "step": 57950 |
| }, |
| { |
| "epoch": 6.242600365945538, |
| "grad_norm": 0.7320728898048401, |
| "learning_rate": 0.00022584204288330996, |
| "loss": 3.2673, |
| "step": 58000 |
| }, |
| { |
| "epoch": 6.242600365945538, |
| "eval_accuracy": 0.3850747439513209, |
| "eval_loss": 3.380539894104004, |
| "eval_runtime": 179.9128, |
| "eval_samples_per_second": 100.11, |
| "eval_steps_per_second": 6.259, |
| "step": 58000 |
| }, |
| { |
| "epoch": 6.247981917985147, |
| "grad_norm": 0.7100010514259338, |
| "learning_rate": 0.00022551880185324853, |
| "loss": 3.2534, |
| "step": 58050 |
| }, |
| { |
| "epoch": 6.253363470024755, |
| "grad_norm": 0.6797668933868408, |
| "learning_rate": 0.00022519556082318715, |
| "loss": 3.2727, |
| "step": 58100 |
| }, |
| { |
| "epoch": 6.258745022064363, |
| "grad_norm": 0.7310889959335327, |
| "learning_rate": 0.00022487231979312572, |
| "loss": 3.2822, |
| "step": 58150 |
| }, |
| { |
| "epoch": 6.264126574103972, |
| "grad_norm": 0.7226874232292175, |
| "learning_rate": 0.0002245490787630643, |
| "loss": 3.2832, |
| "step": 58200 |
| }, |
| { |
| "epoch": 6.26950812614358, |
| "grad_norm": 0.7260108590126038, |
| "learning_rate": 0.0002242258377330029, |
| "loss": 3.2615, |
| "step": 58250 |
| }, |
| { |
| "epoch": 6.274889678183188, |
| "grad_norm": 0.6836520433425903, |
| "learning_rate": 0.00022390259670294148, |
| "loss": 3.2678, |
| "step": 58300 |
| }, |
| { |
| "epoch": 6.280271230222796, |
| "grad_norm": 0.7636072039604187, |
| "learning_rate": 0.00022357935567288005, |
| "loss": 3.27, |
| "step": 58350 |
| }, |
| { |
| "epoch": 6.285652782262405, |
| "grad_norm": 0.7172989845275879, |
| "learning_rate": 0.00022325611464281864, |
| "loss": 3.2646, |
| "step": 58400 |
| }, |
| { |
| "epoch": 6.2910343343020125, |
| "grad_norm": 0.7145113945007324, |
| "learning_rate": 0.00022293287361275723, |
| "loss": 3.2717, |
| "step": 58450 |
| }, |
| { |
| "epoch": 6.296415886341621, |
| "grad_norm": 0.7093191146850586, |
| "learning_rate": 0.0002226096325826958, |
| "loss": 3.2584, |
| "step": 58500 |
| }, |
| { |
| "epoch": 6.301797438381229, |
| "grad_norm": 0.6939222812652588, |
| "learning_rate": 0.0002222863915526344, |
| "loss": 3.2631, |
| "step": 58550 |
| }, |
| { |
| "epoch": 6.307178990420837, |
| "grad_norm": 0.7617230415344238, |
| "learning_rate": 0.00022196315052257296, |
| "loss": 3.2769, |
| "step": 58600 |
| }, |
| { |
| "epoch": 6.312560542460446, |
| "grad_norm": 0.694742739200592, |
| "learning_rate": 0.00022163990949251156, |
| "loss": 3.2843, |
| "step": 58650 |
| }, |
| { |
| "epoch": 6.3179420945000535, |
| "grad_norm": 0.6639745235443115, |
| "learning_rate": 0.00022131666846245015, |
| "loss": 3.2728, |
| "step": 58700 |
| }, |
| { |
| "epoch": 6.323323646539662, |
| "grad_norm": 0.7379493117332458, |
| "learning_rate": 0.00022099342743238872, |
| "loss": 3.2663, |
| "step": 58750 |
| }, |
| { |
| "epoch": 6.328705198579271, |
| "grad_norm": 0.7672418355941772, |
| "learning_rate": 0.00022067018640232734, |
| "loss": 3.2839, |
| "step": 58800 |
| }, |
| { |
| "epoch": 6.334086750618878, |
| "grad_norm": 0.7370325922966003, |
| "learning_rate": 0.0002203469453722659, |
| "loss": 3.2852, |
| "step": 58850 |
| }, |
| { |
| "epoch": 6.339468302658487, |
| "grad_norm": 0.7163979411125183, |
| "learning_rate": 0.00022002370434220448, |
| "loss": 3.2848, |
| "step": 58900 |
| }, |
| { |
| "epoch": 6.344849854698095, |
| "grad_norm": 0.7386805415153503, |
| "learning_rate": 0.00021970046331214307, |
| "loss": 3.2539, |
| "step": 58950 |
| }, |
| { |
| "epoch": 6.350231406737703, |
| "grad_norm": 0.7379843592643738, |
| "learning_rate": 0.00021937722228208167, |
| "loss": 3.2727, |
| "step": 59000 |
| }, |
| { |
| "epoch": 6.350231406737703, |
| "eval_accuracy": 0.3851716623272514, |
| "eval_loss": 3.3763844966888428, |
| "eval_runtime": 180.0816, |
| "eval_samples_per_second": 100.016, |
| "eval_steps_per_second": 6.253, |
| "step": 59000 |
| }, |
| { |
| "epoch": 6.355612958777312, |
| "grad_norm": 0.6990553736686707, |
| "learning_rate": 0.00021905398125202024, |
| "loss": 3.2635, |
| "step": 59050 |
| }, |
| { |
| "epoch": 6.360994510816919, |
| "grad_norm": 0.7014365792274475, |
| "learning_rate": 0.00021873074022195883, |
| "loss": 3.2834, |
| "step": 59100 |
| }, |
| { |
| "epoch": 6.366376062856528, |
| "grad_norm": 0.6956151127815247, |
| "learning_rate": 0.0002184074991918974, |
| "loss": 3.2644, |
| "step": 59150 |
| }, |
| { |
| "epoch": 6.371757614896136, |
| "grad_norm": 0.7179141640663147, |
| "learning_rate": 0.00021808425816183597, |
| "loss": 3.2749, |
| "step": 59200 |
| }, |
| { |
| "epoch": 6.377139166935744, |
| "grad_norm": 0.7074273824691772, |
| "learning_rate": 0.0002177610171317746, |
| "loss": 3.2607, |
| "step": 59250 |
| }, |
| { |
| "epoch": 6.382520718975353, |
| "grad_norm": 0.6928147673606873, |
| "learning_rate": 0.00021743777610171315, |
| "loss": 3.2804, |
| "step": 59300 |
| }, |
| { |
| "epoch": 6.387902271014961, |
| "grad_norm": 0.7080892324447632, |
| "learning_rate": 0.00021711453507165172, |
| "loss": 3.2897, |
| "step": 59350 |
| }, |
| { |
| "epoch": 6.393283823054569, |
| "grad_norm": 0.7480721473693848, |
| "learning_rate": 0.00021679129404159034, |
| "loss": 3.275, |
| "step": 59400 |
| }, |
| { |
| "epoch": 6.398665375094177, |
| "grad_norm": 0.7683929800987244, |
| "learning_rate": 0.0002164680530115289, |
| "loss": 3.2511, |
| "step": 59450 |
| }, |
| { |
| "epoch": 6.404046927133785, |
| "grad_norm": 0.7302149534225464, |
| "learning_rate": 0.00021614481198146748, |
| "loss": 3.2587, |
| "step": 59500 |
| }, |
| { |
| "epoch": 6.4094284791733935, |
| "grad_norm": 0.7285591959953308, |
| "learning_rate": 0.00021582157095140607, |
| "loss": 3.2727, |
| "step": 59550 |
| }, |
| { |
| "epoch": 6.414810031213002, |
| "grad_norm": 0.7607294917106628, |
| "learning_rate": 0.00021549832992134467, |
| "loss": 3.2927, |
| "step": 59600 |
| }, |
| { |
| "epoch": 6.42019158325261, |
| "grad_norm": 0.7400927543640137, |
| "learning_rate": 0.00021517508889128324, |
| "loss": 3.2775, |
| "step": 59650 |
| }, |
| { |
| "epoch": 6.425573135292218, |
| "grad_norm": 0.7432016730308533, |
| "learning_rate": 0.00021485184786122183, |
| "loss": 3.2883, |
| "step": 59700 |
| }, |
| { |
| "epoch": 6.430954687331827, |
| "grad_norm": 0.754863977432251, |
| "learning_rate": 0.0002145286068311604, |
| "loss": 3.2949, |
| "step": 59750 |
| }, |
| { |
| "epoch": 6.4363362393714345, |
| "grad_norm": 0.7521831393241882, |
| "learning_rate": 0.00021420536580109902, |
| "loss": 3.2847, |
| "step": 59800 |
| }, |
| { |
| "epoch": 6.441717791411043, |
| "grad_norm": 0.6796225309371948, |
| "learning_rate": 0.0002138821247710376, |
| "loss": 3.2842, |
| "step": 59850 |
| }, |
| { |
| "epoch": 6.447099343450651, |
| "grad_norm": 0.7475155591964722, |
| "learning_rate": 0.00021355888374097615, |
| "loss": 3.2863, |
| "step": 59900 |
| }, |
| { |
| "epoch": 6.452480895490259, |
| "grad_norm": 0.7627286314964294, |
| "learning_rate": 0.00021323564271091478, |
| "loss": 3.2928, |
| "step": 59950 |
| }, |
| { |
| "epoch": 6.457862447529868, |
| "grad_norm": 0.7810362577438354, |
| "learning_rate": 0.00021291240168085334, |
| "loss": 3.2923, |
| "step": 60000 |
| }, |
| { |
| "epoch": 6.457862447529868, |
| "eval_accuracy": 0.38576631958226004, |
| "eval_loss": 3.371659278869629, |
| "eval_runtime": 179.7476, |
| "eval_samples_per_second": 100.202, |
| "eval_steps_per_second": 6.264, |
| "step": 60000 |
| }, |
| { |
| "epoch": 6.4632439995694755, |
| "grad_norm": 0.731270968914032, |
| "learning_rate": 0.0002125891606507919, |
| "loss": 3.2814, |
| "step": 60050 |
| }, |
| { |
| "epoch": 6.468625551609084, |
| "grad_norm": 0.7113689184188843, |
| "learning_rate": 0.0002122659196207305, |
| "loss": 3.2691, |
| "step": 60100 |
| }, |
| { |
| "epoch": 6.474007103648693, |
| "grad_norm": 0.7589284777641296, |
| "learning_rate": 0.0002119426785906691, |
| "loss": 3.2758, |
| "step": 60150 |
| }, |
| { |
| "epoch": 6.4793886556883, |
| "grad_norm": 0.7309033870697021, |
| "learning_rate": 0.00021161943756060767, |
| "loss": 3.2725, |
| "step": 60200 |
| }, |
| { |
| "epoch": 6.484770207727909, |
| "grad_norm": 0.7444686889648438, |
| "learning_rate": 0.00021130266135114747, |
| "loss": 3.2851, |
| "step": 60250 |
| }, |
| { |
| "epoch": 6.490151759767517, |
| "grad_norm": 0.7868755459785461, |
| "learning_rate": 0.00021097942032108607, |
| "loss": 3.2974, |
| "step": 60300 |
| }, |
| { |
| "epoch": 6.495533311807125, |
| "grad_norm": 0.720824122428894, |
| "learning_rate": 0.00021065617929102466, |
| "loss": 3.2745, |
| "step": 60350 |
| }, |
| { |
| "epoch": 6.500914863846734, |
| "grad_norm": 0.717765748500824, |
| "learning_rate": 0.00021033293826096323, |
| "loss": 3.2889, |
| "step": 60400 |
| }, |
| { |
| "epoch": 6.506296415886341, |
| "grad_norm": 0.7933386564254761, |
| "learning_rate": 0.0002100096972309018, |
| "loss": 3.2683, |
| "step": 60450 |
| }, |
| { |
| "epoch": 6.51167796792595, |
| "grad_norm": 0.7101936936378479, |
| "learning_rate": 0.00020968645620084042, |
| "loss": 3.2728, |
| "step": 60500 |
| }, |
| { |
| "epoch": 6.517059519965558, |
| "grad_norm": 0.6985874176025391, |
| "learning_rate": 0.000209363215170779, |
| "loss": 3.2901, |
| "step": 60550 |
| }, |
| { |
| "epoch": 6.522441072005166, |
| "grad_norm": 0.7263524532318115, |
| "learning_rate": 0.00020903997414071758, |
| "loss": 3.2667, |
| "step": 60600 |
| }, |
| { |
| "epoch": 6.5278226240447745, |
| "grad_norm": 0.7178045511245728, |
| "learning_rate": 0.00020871673311065618, |
| "loss": 3.282, |
| "step": 60650 |
| }, |
| { |
| "epoch": 6.533204176084383, |
| "grad_norm": 0.7246670126914978, |
| "learning_rate": 0.00020839349208059475, |
| "loss": 3.2883, |
| "step": 60700 |
| }, |
| { |
| "epoch": 6.538585728123991, |
| "grad_norm": 0.7198054194450378, |
| "learning_rate": 0.00020807025105053334, |
| "loss": 3.2974, |
| "step": 60750 |
| }, |
| { |
| "epoch": 6.543967280163599, |
| "grad_norm": 0.7221017479896545, |
| "learning_rate": 0.0002077470100204719, |
| "loss": 3.2856, |
| "step": 60800 |
| }, |
| { |
| "epoch": 6.549348832203208, |
| "grad_norm": 0.7302899360656738, |
| "learning_rate": 0.00020742376899041048, |
| "loss": 3.2856, |
| "step": 60850 |
| }, |
| { |
| "epoch": 6.5547303842428155, |
| "grad_norm": 0.7253843545913696, |
| "learning_rate": 0.0002071005279603491, |
| "loss": 3.2752, |
| "step": 60900 |
| }, |
| { |
| "epoch": 6.560111936282424, |
| "grad_norm": 0.7374371886253357, |
| "learning_rate": 0.00020677728693028766, |
| "loss": 3.2728, |
| "step": 60950 |
| }, |
| { |
| "epoch": 6.565493488322032, |
| "grad_norm": 0.70008385181427, |
| "learning_rate": 0.00020645404590022623, |
| "loss": 3.2689, |
| "step": 61000 |
| }, |
| { |
| "epoch": 6.565493488322032, |
| "eval_accuracy": 0.3859744985153126, |
| "eval_loss": 3.3697760105133057, |
| "eval_runtime": 180.5619, |
| "eval_samples_per_second": 99.75, |
| "eval_steps_per_second": 6.236, |
| "step": 61000 |
| }, |
| { |
| "epoch": 6.57087504036164, |
| "grad_norm": 0.7110406160354614, |
| "learning_rate": 0.00020613080487016485, |
| "loss": 3.289, |
| "step": 61050 |
| }, |
| { |
| "epoch": 6.576256592401249, |
| "grad_norm": 0.8671207427978516, |
| "learning_rate": 0.00020580756384010342, |
| "loss": 3.2676, |
| "step": 61100 |
| }, |
| { |
| "epoch": 6.5816381444408565, |
| "grad_norm": 0.797752320766449, |
| "learning_rate": 0.000205484322810042, |
| "loss": 3.2827, |
| "step": 61150 |
| }, |
| { |
| "epoch": 6.587019696480465, |
| "grad_norm": 0.7081204652786255, |
| "learning_rate": 0.00020516108177998058, |
| "loss": 3.2917, |
| "step": 61200 |
| }, |
| { |
| "epoch": 6.592401248520073, |
| "grad_norm": 0.7679519057273865, |
| "learning_rate": 0.00020483784074991918, |
| "loss": 3.2761, |
| "step": 61250 |
| }, |
| { |
| "epoch": 6.597782800559681, |
| "grad_norm": 0.7073348164558411, |
| "learning_rate": 0.00020451459971985775, |
| "loss": 3.2917, |
| "step": 61300 |
| }, |
| { |
| "epoch": 6.60316435259929, |
| "grad_norm": 0.714297354221344, |
| "learning_rate": 0.00020419135868979634, |
| "loss": 3.2842, |
| "step": 61350 |
| }, |
| { |
| "epoch": 6.608545904638898, |
| "grad_norm": 0.7196522355079651, |
| "learning_rate": 0.0002038681176597349, |
| "loss": 3.2688, |
| "step": 61400 |
| }, |
| { |
| "epoch": 6.613927456678506, |
| "grad_norm": 0.7500666975975037, |
| "learning_rate": 0.0002035448766296735, |
| "loss": 3.2923, |
| "step": 61450 |
| }, |
| { |
| "epoch": 6.619309008718115, |
| "grad_norm": 0.7291411757469177, |
| "learning_rate": 0.0002032216355996121, |
| "loss": 3.2894, |
| "step": 61500 |
| }, |
| { |
| "epoch": 6.624690560757722, |
| "grad_norm": 0.7208287119865417, |
| "learning_rate": 0.00020289839456955067, |
| "loss": 3.3013, |
| "step": 61550 |
| }, |
| { |
| "epoch": 6.630072112797331, |
| "grad_norm": 0.7192732691764832, |
| "learning_rate": 0.0002025751535394893, |
| "loss": 3.2926, |
| "step": 61600 |
| }, |
| { |
| "epoch": 6.635453664836939, |
| "grad_norm": 0.769903302192688, |
| "learning_rate": 0.00020225191250942785, |
| "loss": 3.2713, |
| "step": 61650 |
| }, |
| { |
| "epoch": 6.640835216876547, |
| "grad_norm": 0.758073091506958, |
| "learning_rate": 0.00020192867147936642, |
| "loss": 3.2694, |
| "step": 61700 |
| }, |
| { |
| "epoch": 6.6462167689161555, |
| "grad_norm": 0.7105669379234314, |
| "learning_rate": 0.00020160543044930502, |
| "loss": 3.2889, |
| "step": 61750 |
| }, |
| { |
| "epoch": 6.651598320955763, |
| "grad_norm": 0.7682111263275146, |
| "learning_rate": 0.0002012821894192436, |
| "loss": 3.2897, |
| "step": 61800 |
| }, |
| { |
| "epoch": 6.656979872995372, |
| "grad_norm": 0.7516341209411621, |
| "learning_rate": 0.00020095894838918218, |
| "loss": 3.2793, |
| "step": 61850 |
| }, |
| { |
| "epoch": 6.66236142503498, |
| "grad_norm": 0.7740298509597778, |
| "learning_rate": 0.00020063570735912077, |
| "loss": 3.2811, |
| "step": 61900 |
| }, |
| { |
| "epoch": 6.667742977074588, |
| "grad_norm": 0.7486993074417114, |
| "learning_rate": 0.00020031246632905934, |
| "loss": 3.2811, |
| "step": 61950 |
| }, |
| { |
| "epoch": 6.6731245291141965, |
| "grad_norm": 0.8277780413627625, |
| "learning_rate": 0.00019998922529899794, |
| "loss": 3.2853, |
| "step": 62000 |
| }, |
| { |
| "epoch": 6.6731245291141965, |
| "eval_accuracy": 0.3862082108770016, |
| "eval_loss": 3.3669943809509277, |
| "eval_runtime": 179.7428, |
| "eval_samples_per_second": 100.204, |
| "eval_steps_per_second": 6.265, |
| "step": 62000 |
| }, |
| { |
| "epoch": 6.678506081153805, |
| "grad_norm": 0.7306733727455139, |
| "learning_rate": 0.00019966598426893653, |
| "loss": 3.2939, |
| "step": 62050 |
| }, |
| { |
| "epoch": 6.683887633193413, |
| "grad_norm": 0.7362749576568604, |
| "learning_rate": 0.0001993427432388751, |
| "loss": 3.2821, |
| "step": 62100 |
| }, |
| { |
| "epoch": 6.689269185233021, |
| "grad_norm": 0.7791106700897217, |
| "learning_rate": 0.00019901950220881367, |
| "loss": 3.2954, |
| "step": 62150 |
| }, |
| { |
| "epoch": 6.69465073727263, |
| "grad_norm": 0.7366771101951599, |
| "learning_rate": 0.0001986962611787523, |
| "loss": 3.2765, |
| "step": 62200 |
| }, |
| { |
| "epoch": 6.7000322893122375, |
| "grad_norm": 0.7491232752799988, |
| "learning_rate": 0.00019837302014869085, |
| "loss": 3.301, |
| "step": 62250 |
| }, |
| { |
| "epoch": 6.705413841351846, |
| "grad_norm": 0.742691695690155, |
| "learning_rate": 0.00019804977911862942, |
| "loss": 3.2811, |
| "step": 62300 |
| }, |
| { |
| "epoch": 6.710795393391454, |
| "grad_norm": 0.7395520210266113, |
| "learning_rate": 0.00019772653808856804, |
| "loss": 3.2865, |
| "step": 62350 |
| }, |
| { |
| "epoch": 6.716176945431062, |
| "grad_norm": 0.7324630618095398, |
| "learning_rate": 0.0001974032970585066, |
| "loss": 3.2892, |
| "step": 62400 |
| }, |
| { |
| "epoch": 6.721558497470671, |
| "grad_norm": 0.7284770011901855, |
| "learning_rate": 0.00019708005602844518, |
| "loss": 3.3008, |
| "step": 62450 |
| }, |
| { |
| "epoch": 6.7269400495102785, |
| "grad_norm": 0.7420316338539124, |
| "learning_rate": 0.00019675681499838377, |
| "loss": 3.2911, |
| "step": 62500 |
| }, |
| { |
| "epoch": 6.732321601549887, |
| "grad_norm": 0.7340070605278015, |
| "learning_rate": 0.00019643357396832237, |
| "loss": 3.2747, |
| "step": 62550 |
| }, |
| { |
| "epoch": 6.737703153589496, |
| "grad_norm": 0.7871265411376953, |
| "learning_rate": 0.00019611033293826096, |
| "loss": 3.297, |
| "step": 62600 |
| }, |
| { |
| "epoch": 6.743084705629103, |
| "grad_norm": 0.7386654615402222, |
| "learning_rate": 0.00019578709190819953, |
| "loss": 3.2891, |
| "step": 62650 |
| }, |
| { |
| "epoch": 6.748466257668712, |
| "grad_norm": 0.7021500468254089, |
| "learning_rate": 0.0001954638508781381, |
| "loss": 3.2665, |
| "step": 62700 |
| }, |
| { |
| "epoch": 6.75384780970832, |
| "grad_norm": 0.7563971281051636, |
| "learning_rate": 0.00019514060984807672, |
| "loss": 3.2821, |
| "step": 62750 |
| }, |
| { |
| "epoch": 6.759229361747928, |
| "grad_norm": 0.7392533421516418, |
| "learning_rate": 0.0001948173688180153, |
| "loss": 3.2866, |
| "step": 62800 |
| }, |
| { |
| "epoch": 6.7646109137875365, |
| "grad_norm": 0.7882697582244873, |
| "learning_rate": 0.00019449412778795386, |
| "loss": 3.3085, |
| "step": 62850 |
| }, |
| { |
| "epoch": 6.769992465827144, |
| "grad_norm": 0.751366138458252, |
| "learning_rate": 0.00019417088675789245, |
| "loss": 3.2849, |
| "step": 62900 |
| }, |
| { |
| "epoch": 6.775374017866753, |
| "grad_norm": 0.7440650463104248, |
| "learning_rate": 0.00019384764572783104, |
| "loss": 3.2891, |
| "step": 62950 |
| }, |
| { |
| "epoch": 6.780755569906361, |
| "grad_norm": 0.7576360702514648, |
| "learning_rate": 0.0001935244046977696, |
| "loss": 3.2815, |
| "step": 63000 |
| }, |
| { |
| "epoch": 6.780755569906361, |
| "eval_accuracy": 0.3868667560300451, |
| "eval_loss": 3.360293388366699, |
| "eval_runtime": 180.0715, |
| "eval_samples_per_second": 100.021, |
| "eval_steps_per_second": 6.253, |
| "step": 63000 |
| }, |
| { |
| "epoch": 6.786137121945969, |
| "grad_norm": 0.7954241633415222, |
| "learning_rate": 0.0001932011636677082, |
| "loss": 3.2861, |
| "step": 63050 |
| }, |
| { |
| "epoch": 6.7915186739855775, |
| "grad_norm": 0.7142677307128906, |
| "learning_rate": 0.00019287792263764677, |
| "loss": 3.2811, |
| "step": 63100 |
| }, |
| { |
| "epoch": 6.796900226025185, |
| "grad_norm": 0.7331296801567078, |
| "learning_rate": 0.00019255468160758537, |
| "loss": 3.2915, |
| "step": 63150 |
| }, |
| { |
| "epoch": 6.802281778064794, |
| "grad_norm": 0.7577782273292542, |
| "learning_rate": 0.00019223144057752396, |
| "loss": 3.2868, |
| "step": 63200 |
| }, |
| { |
| "epoch": 6.807663330104402, |
| "grad_norm": 0.7805274128913879, |
| "learning_rate": 0.00019190819954746253, |
| "loss": 3.2869, |
| "step": 63250 |
| }, |
| { |
| "epoch": 6.813044882144011, |
| "grad_norm": 0.7110979557037354, |
| "learning_rate": 0.0001915849585174011, |
| "loss": 3.2906, |
| "step": 63300 |
| }, |
| { |
| "epoch": 6.8184264341836185, |
| "grad_norm": 0.7497155070304871, |
| "learning_rate": 0.00019126171748733972, |
| "loss": 3.2839, |
| "step": 63350 |
| }, |
| { |
| "epoch": 6.823807986223227, |
| "grad_norm": 0.7474818825721741, |
| "learning_rate": 0.0001909384764572783, |
| "loss": 3.277, |
| "step": 63400 |
| }, |
| { |
| "epoch": 6.829189538262835, |
| "grad_norm": 0.7573341727256775, |
| "learning_rate": 0.00019061523542721688, |
| "loss": 3.2829, |
| "step": 63450 |
| }, |
| { |
| "epoch": 6.834571090302443, |
| "grad_norm": 0.6964619755744934, |
| "learning_rate": 0.00019029199439715548, |
| "loss": 3.2674, |
| "step": 63500 |
| }, |
| { |
| "epoch": 6.839952642342052, |
| "grad_norm": 0.7417646646499634, |
| "learning_rate": 0.00018996875336709404, |
| "loss": 3.3016, |
| "step": 63550 |
| }, |
| { |
| "epoch": 6.8453341943816595, |
| "grad_norm": 0.7879679203033447, |
| "learning_rate": 0.00018964551233703264, |
| "loss": 3.2684, |
| "step": 63600 |
| }, |
| { |
| "epoch": 6.850715746421268, |
| "grad_norm": 0.7311843633651733, |
| "learning_rate": 0.0001893222713069712, |
| "loss": 3.286, |
| "step": 63650 |
| }, |
| { |
| "epoch": 6.856097298460876, |
| "grad_norm": 0.7371758222579956, |
| "learning_rate": 0.0001889990302769098, |
| "loss": 3.2775, |
| "step": 63700 |
| }, |
| { |
| "epoch": 6.861478850500484, |
| "grad_norm": 0.7263456583023071, |
| "learning_rate": 0.0001886757892468484, |
| "loss": 3.2652, |
| "step": 63750 |
| }, |
| { |
| "epoch": 6.866860402540093, |
| "grad_norm": 0.7316994071006775, |
| "learning_rate": 0.00018835254821678696, |
| "loss": 3.2727, |
| "step": 63800 |
| }, |
| { |
| "epoch": 6.8722419545797, |
| "grad_norm": 0.7389889359474182, |
| "learning_rate": 0.00018802930718672553, |
| "loss": 3.3056, |
| "step": 63850 |
| }, |
| { |
| "epoch": 6.877623506619309, |
| "grad_norm": 0.7888100743293762, |
| "learning_rate": 0.00018770606615666415, |
| "loss": 3.2851, |
| "step": 63900 |
| }, |
| { |
| "epoch": 6.8830050586589175, |
| "grad_norm": 0.7792270183563232, |
| "learning_rate": 0.00018738282512660272, |
| "loss": 3.2949, |
| "step": 63950 |
| }, |
| { |
| "epoch": 6.888386610698525, |
| "grad_norm": 0.7635626196861267, |
| "learning_rate": 0.0001870595840965413, |
| "loss": 3.2878, |
| "step": 64000 |
| }, |
| { |
| "epoch": 6.888386610698525, |
| "eval_accuracy": 0.3867621232990798, |
| "eval_loss": 3.3606150150299072, |
| "eval_runtime": 180.6437, |
| "eval_samples_per_second": 99.705, |
| "eval_steps_per_second": 6.233, |
| "step": 64000 |
| }, |
| { |
| "epoch": 6.893768162738134, |
| "grad_norm": 0.7803065180778503, |
| "learning_rate": 0.0001867363430664799, |
| "loss": 3.2846, |
| "step": 64050 |
| }, |
| { |
| "epoch": 6.899149714777742, |
| "grad_norm": 0.7100071310997009, |
| "learning_rate": 0.00018641310203641848, |
| "loss": 3.3034, |
| "step": 64100 |
| }, |
| { |
| "epoch": 6.90453126681735, |
| "grad_norm": 0.7439180016517639, |
| "learning_rate": 0.00018608986100635705, |
| "loss": 3.2756, |
| "step": 64150 |
| }, |
| { |
| "epoch": 6.9099128188569585, |
| "grad_norm": 0.7899887561798096, |
| "learning_rate": 0.00018576661997629564, |
| "loss": 3.2778, |
| "step": 64200 |
| }, |
| { |
| "epoch": 6.915294370896566, |
| "grad_norm": 0.7477514147758484, |
| "learning_rate": 0.00018544337894623423, |
| "loss": 3.2992, |
| "step": 64250 |
| }, |
| { |
| "epoch": 6.920675922936175, |
| "grad_norm": 0.7305505275726318, |
| "learning_rate": 0.00018512660273677404, |
| "loss": 3.2864, |
| "step": 64300 |
| }, |
| { |
| "epoch": 6.926057474975783, |
| "grad_norm": 0.7840530872344971, |
| "learning_rate": 0.0001848033617067126, |
| "loss": 3.2804, |
| "step": 64350 |
| }, |
| { |
| "epoch": 6.931439027015391, |
| "grad_norm": 0.7968816161155701, |
| "learning_rate": 0.00018448012067665123, |
| "loss": 3.2859, |
| "step": 64400 |
| }, |
| { |
| "epoch": 6.9368205790549995, |
| "grad_norm": 0.7263173460960388, |
| "learning_rate": 0.0001841568796465898, |
| "loss": 3.2938, |
| "step": 64450 |
| }, |
| { |
| "epoch": 6.942202131094608, |
| "grad_norm": 0.74007248878479, |
| "learning_rate": 0.00018383363861652837, |
| "loss": 3.2647, |
| "step": 64500 |
| }, |
| { |
| "epoch": 6.947583683134216, |
| "grad_norm": 0.730910062789917, |
| "learning_rate": 0.00018351039758646696, |
| "loss": 3.2913, |
| "step": 64550 |
| }, |
| { |
| "epoch": 6.952965235173824, |
| "grad_norm": 0.7842426300048828, |
| "learning_rate": 0.00018318715655640555, |
| "loss": 3.2938, |
| "step": 64600 |
| }, |
| { |
| "epoch": 6.958346787213433, |
| "grad_norm": 0.7127736210823059, |
| "learning_rate": 0.00018286391552634412, |
| "loss": 3.2837, |
| "step": 64650 |
| }, |
| { |
| "epoch": 6.9637283392530405, |
| "grad_norm": 0.7509408593177795, |
| "learning_rate": 0.00018254067449628272, |
| "loss": 3.2829, |
| "step": 64700 |
| }, |
| { |
| "epoch": 6.969109891292649, |
| "grad_norm": 0.7198226451873779, |
| "learning_rate": 0.00018221743346622128, |
| "loss": 3.2862, |
| "step": 64750 |
| }, |
| { |
| "epoch": 6.974491443332257, |
| "grad_norm": 0.7937029600143433, |
| "learning_rate": 0.00018189419243615988, |
| "loss": 3.2979, |
| "step": 64800 |
| }, |
| { |
| "epoch": 6.979872995371865, |
| "grad_norm": 0.758535623550415, |
| "learning_rate": 0.00018157095140609847, |
| "loss": 3.2594, |
| "step": 64850 |
| }, |
| { |
| "epoch": 6.985254547411474, |
| "grad_norm": 0.8164926767349243, |
| "learning_rate": 0.00018124771037603704, |
| "loss": 3.2843, |
| "step": 64900 |
| }, |
| { |
| "epoch": 6.990636099451081, |
| "grad_norm": 0.7364434003829956, |
| "learning_rate": 0.0001809244693459756, |
| "loss": 3.2898, |
| "step": 64950 |
| }, |
| { |
| "epoch": 6.99601765149069, |
| "grad_norm": 0.7529191970825195, |
| "learning_rate": 0.00018060122831591423, |
| "loss": 3.2674, |
| "step": 65000 |
| }, |
| { |
| "epoch": 6.99601765149069, |
| "eval_accuracy": 0.3877488001733231, |
| "eval_loss": 3.354095458984375, |
| "eval_runtime": 180.0368, |
| "eval_samples_per_second": 100.041, |
| "eval_steps_per_second": 6.254, |
| "step": 65000 |
| }, |
| { |
| "epoch": 7.0013992035302985, |
| "grad_norm": 0.7346994876861572, |
| "learning_rate": 0.0001802779872858528, |
| "loss": 3.2675, |
| "step": 65050 |
| }, |
| { |
| "epoch": 7.006780755569906, |
| "grad_norm": 0.7564079165458679, |
| "learning_rate": 0.00017995474625579137, |
| "loss": 3.2023, |
| "step": 65100 |
| }, |
| { |
| "epoch": 7.012162307609515, |
| "grad_norm": 0.7516438364982605, |
| "learning_rate": 0.00017963150522573, |
| "loss": 3.2093, |
| "step": 65150 |
| }, |
| { |
| "epoch": 7.017543859649122, |
| "grad_norm": 0.757781445980072, |
| "learning_rate": 0.00017930826419566856, |
| "loss": 3.2116, |
| "step": 65200 |
| }, |
| { |
| "epoch": 7.022925411688731, |
| "grad_norm": 0.7500360608100891, |
| "learning_rate": 0.00017898502316560712, |
| "loss": 3.2007, |
| "step": 65250 |
| }, |
| { |
| "epoch": 7.0283069637283395, |
| "grad_norm": 0.7233052849769592, |
| "learning_rate": 0.00017866178213554572, |
| "loss": 3.2189, |
| "step": 65300 |
| }, |
| { |
| "epoch": 7.033688515767947, |
| "grad_norm": 0.8041335940361023, |
| "learning_rate": 0.0001783385411054843, |
| "loss": 3.1901, |
| "step": 65350 |
| }, |
| { |
| "epoch": 7.039070067807556, |
| "grad_norm": 0.7104873657226562, |
| "learning_rate": 0.0001780153000754229, |
| "loss": 3.1969, |
| "step": 65400 |
| }, |
| { |
| "epoch": 7.044451619847164, |
| "grad_norm": 0.7547011375427246, |
| "learning_rate": 0.00017769205904536147, |
| "loss": 3.2141, |
| "step": 65450 |
| }, |
| { |
| "epoch": 7.049833171886772, |
| "grad_norm": 0.7907810211181641, |
| "learning_rate": 0.00017736881801530004, |
| "loss": 3.2051, |
| "step": 65500 |
| }, |
| { |
| "epoch": 7.0552147239263805, |
| "grad_norm": 0.7557851076126099, |
| "learning_rate": 0.00017704557698523866, |
| "loss": 3.209, |
| "step": 65550 |
| }, |
| { |
| "epoch": 7.060596275965988, |
| "grad_norm": 0.7390865087509155, |
| "learning_rate": 0.00017672233595517723, |
| "loss": 3.2206, |
| "step": 65600 |
| }, |
| { |
| "epoch": 7.065977828005597, |
| "grad_norm": 0.7852340340614319, |
| "learning_rate": 0.0001763990949251158, |
| "loss": 3.2176, |
| "step": 65650 |
| }, |
| { |
| "epoch": 7.071359380045205, |
| "grad_norm": 0.7627012729644775, |
| "learning_rate": 0.00017607585389505442, |
| "loss": 3.2183, |
| "step": 65700 |
| }, |
| { |
| "epoch": 7.076740932084813, |
| "grad_norm": 0.7575951814651489, |
| "learning_rate": 0.000175752612864993, |
| "loss": 3.2181, |
| "step": 65750 |
| }, |
| { |
| "epoch": 7.0821224841244215, |
| "grad_norm": 0.7672650218009949, |
| "learning_rate": 0.00017542937183493156, |
| "loss": 3.2102, |
| "step": 65800 |
| }, |
| { |
| "epoch": 7.08750403616403, |
| "grad_norm": 0.7804780006408691, |
| "learning_rate": 0.00017510613080487015, |
| "loss": 3.2121, |
| "step": 65850 |
| }, |
| { |
| "epoch": 7.092885588203638, |
| "grad_norm": 0.7676131725311279, |
| "learning_rate": 0.00017478288977480874, |
| "loss": 3.2017, |
| "step": 65900 |
| }, |
| { |
| "epoch": 7.098267140243246, |
| "grad_norm": 0.7814217805862427, |
| "learning_rate": 0.00017446611356534855, |
| "loss": 3.2032, |
| "step": 65950 |
| }, |
| { |
| "epoch": 7.103648692282855, |
| "grad_norm": 0.7216848731040955, |
| "learning_rate": 0.00017414287253528712, |
| "loss": 3.2193, |
| "step": 66000 |
| }, |
| { |
| "epoch": 7.103648692282855, |
| "eval_accuracy": 0.38735189117413676, |
| "eval_loss": 3.357109785079956, |
| "eval_runtime": 179.9063, |
| "eval_samples_per_second": 100.113, |
| "eval_steps_per_second": 6.259, |
| "step": 66000 |
| }, |
| { |
| "epoch": 7.109030244322462, |
| "grad_norm": 0.754563570022583, |
| "learning_rate": 0.00017381963150522569, |
| "loss": 3.2194, |
| "step": 66050 |
| }, |
| { |
| "epoch": 7.114411796362071, |
| "grad_norm": 0.7799685597419739, |
| "learning_rate": 0.0001734963904751643, |
| "loss": 3.2178, |
| "step": 66100 |
| }, |
| { |
| "epoch": 7.119793348401679, |
| "grad_norm": 0.7888051271438599, |
| "learning_rate": 0.00017317314944510288, |
| "loss": 3.2021, |
| "step": 66150 |
| }, |
| { |
| "epoch": 7.125174900441287, |
| "grad_norm": 0.7203174829483032, |
| "learning_rate": 0.00017284990841504147, |
| "loss": 3.211, |
| "step": 66200 |
| }, |
| { |
| "epoch": 7.130556452480896, |
| "grad_norm": 0.8019355535507202, |
| "learning_rate": 0.00017252666738498007, |
| "loss": 3.203, |
| "step": 66250 |
| }, |
| { |
| "epoch": 7.135938004520503, |
| "grad_norm": 0.7878087162971497, |
| "learning_rate": 0.00017220342635491863, |
| "loss": 3.2177, |
| "step": 66300 |
| }, |
| { |
| "epoch": 7.141319556560112, |
| "grad_norm": 0.7539237141609192, |
| "learning_rate": 0.00017188018532485723, |
| "loss": 3.2079, |
| "step": 66350 |
| }, |
| { |
| "epoch": 7.1467011085997205, |
| "grad_norm": 0.7446153163909912, |
| "learning_rate": 0.0001715569442947958, |
| "loss": 3.1975, |
| "step": 66400 |
| }, |
| { |
| "epoch": 7.152082660639328, |
| "grad_norm": 0.733919084072113, |
| "learning_rate": 0.0001712337032647344, |
| "loss": 3.243, |
| "step": 66450 |
| }, |
| { |
| "epoch": 7.157464212678937, |
| "grad_norm": 0.806229829788208, |
| "learning_rate": 0.00017091046223467298, |
| "loss": 3.2213, |
| "step": 66500 |
| }, |
| { |
| "epoch": 7.162845764718545, |
| "grad_norm": 0.7871570587158203, |
| "learning_rate": 0.00017058722120461155, |
| "loss": 3.2023, |
| "step": 66550 |
| }, |
| { |
| "epoch": 7.168227316758153, |
| "grad_norm": 0.7633065581321716, |
| "learning_rate": 0.00017026398017455012, |
| "loss": 3.2327, |
| "step": 66600 |
| }, |
| { |
| "epoch": 7.1736088687977615, |
| "grad_norm": 0.7800832986831665, |
| "learning_rate": 0.00016994073914448874, |
| "loss": 3.2351, |
| "step": 66650 |
| }, |
| { |
| "epoch": 7.178990420837369, |
| "grad_norm": 0.7342001795768738, |
| "learning_rate": 0.0001696174981144273, |
| "loss": 3.2212, |
| "step": 66700 |
| }, |
| { |
| "epoch": 7.184371972876978, |
| "grad_norm": 0.7810934782028198, |
| "learning_rate": 0.00016929425708436588, |
| "loss": 3.2014, |
| "step": 66750 |
| }, |
| { |
| "epoch": 7.189753524916586, |
| "grad_norm": 0.7123428583145142, |
| "learning_rate": 0.0001689710160543045, |
| "loss": 3.2019, |
| "step": 66800 |
| }, |
| { |
| "epoch": 7.195135076956194, |
| "grad_norm": 0.8075796365737915, |
| "learning_rate": 0.00016864777502424307, |
| "loss": 3.2242, |
| "step": 66850 |
| }, |
| { |
| "epoch": 7.2005166289958025, |
| "grad_norm": 0.8010225892066956, |
| "learning_rate": 0.00016832453399418163, |
| "loss": 3.2183, |
| "step": 66900 |
| }, |
| { |
| "epoch": 7.205898181035411, |
| "grad_norm": 0.7485325336456299, |
| "learning_rate": 0.00016800129296412023, |
| "loss": 3.2022, |
| "step": 66950 |
| }, |
| { |
| "epoch": 7.211279733075019, |
| "grad_norm": 0.8330503702163696, |
| "learning_rate": 0.00016767805193405882, |
| "loss": 3.2274, |
| "step": 67000 |
| }, |
| { |
| "epoch": 7.211279733075019, |
| "eval_accuracy": 0.3875922313619824, |
| "eval_loss": 3.357809066772461, |
| "eval_runtime": 179.9928, |
| "eval_samples_per_second": 100.065, |
| "eval_steps_per_second": 6.256, |
| "step": 67000 |
| }, |
| { |
| "epoch": 7.216661285114627, |
| "grad_norm": 0.7660570740699768, |
| "learning_rate": 0.0001673548109039974, |
| "loss": 3.218, |
| "step": 67050 |
| }, |
| { |
| "epoch": 7.222042837154235, |
| "grad_norm": 0.7876999974250793, |
| "learning_rate": 0.00016703156987393598, |
| "loss": 3.2167, |
| "step": 67100 |
| }, |
| { |
| "epoch": 7.2274243891938434, |
| "grad_norm": 0.7447673082351685, |
| "learning_rate": 0.00016670832884387455, |
| "loss": 3.2343, |
| "step": 67150 |
| }, |
| { |
| "epoch": 7.232805941233452, |
| "grad_norm": 0.7839945554733276, |
| "learning_rate": 0.00016638508781381317, |
| "loss": 3.2155, |
| "step": 67200 |
| }, |
| { |
| "epoch": 7.23818749327306, |
| "grad_norm": 0.7349887490272522, |
| "learning_rate": 0.00016606184678375174, |
| "loss": 3.2318, |
| "step": 67250 |
| }, |
| { |
| "epoch": 7.243569045312668, |
| "grad_norm": 0.7882452011108398, |
| "learning_rate": 0.0001657386057536903, |
| "loss": 3.2308, |
| "step": 67300 |
| }, |
| { |
| "epoch": 7.248950597352277, |
| "grad_norm": 0.7886552810668945, |
| "learning_rate": 0.00016541536472362893, |
| "loss": 3.228, |
| "step": 67350 |
| }, |
| { |
| "epoch": 7.254332149391884, |
| "grad_norm": 0.7716850638389587, |
| "learning_rate": 0.0001650921236935675, |
| "loss": 3.2112, |
| "step": 67400 |
| }, |
| { |
| "epoch": 7.259713701431493, |
| "grad_norm": 0.7802140712738037, |
| "learning_rate": 0.00016476888266350607, |
| "loss": 3.2307, |
| "step": 67450 |
| }, |
| { |
| "epoch": 7.265095253471101, |
| "grad_norm": 0.7871243953704834, |
| "learning_rate": 0.00016444564163344466, |
| "loss": 3.2068, |
| "step": 67500 |
| }, |
| { |
| "epoch": 7.270476805510709, |
| "grad_norm": 0.8147279024124146, |
| "learning_rate": 0.00016412240060338326, |
| "loss": 3.2212, |
| "step": 67550 |
| }, |
| { |
| "epoch": 7.275858357550318, |
| "grad_norm": 0.8272632956504822, |
| "learning_rate": 0.00016379915957332182, |
| "loss": 3.2287, |
| "step": 67600 |
| }, |
| { |
| "epoch": 7.281239909589925, |
| "grad_norm": 0.7926257252693176, |
| "learning_rate": 0.00016347591854326042, |
| "loss": 3.2024, |
| "step": 67650 |
| }, |
| { |
| "epoch": 7.286621461629534, |
| "grad_norm": 0.8117291927337646, |
| "learning_rate": 0.00016315267751319898, |
| "loss": 3.2295, |
| "step": 67700 |
| }, |
| { |
| "epoch": 7.2920030136691425, |
| "grad_norm": 0.7746281623840332, |
| "learning_rate": 0.00016282943648313755, |
| "loss": 3.2416, |
| "step": 67750 |
| }, |
| { |
| "epoch": 7.29738456570875, |
| "grad_norm": 0.7859088182449341, |
| "learning_rate": 0.00016250619545307617, |
| "loss": 3.2179, |
| "step": 67800 |
| }, |
| { |
| "epoch": 7.302766117748359, |
| "grad_norm": 0.749521017074585, |
| "learning_rate": 0.00016218295442301474, |
| "loss": 3.2167, |
| "step": 67850 |
| }, |
| { |
| "epoch": 7.308147669787967, |
| "grad_norm": 0.7174946665763855, |
| "learning_rate": 0.0001618597133929533, |
| "loss": 3.2142, |
| "step": 67900 |
| }, |
| { |
| "epoch": 7.313529221827575, |
| "grad_norm": 0.7507534027099609, |
| "learning_rate": 0.00016153647236289193, |
| "loss": 3.2187, |
| "step": 67950 |
| }, |
| { |
| "epoch": 7.3189107738671835, |
| "grad_norm": 0.7594087719917297, |
| "learning_rate": 0.0001612132313328305, |
| "loss": 3.2201, |
| "step": 68000 |
| }, |
| { |
| "epoch": 7.3189107738671835, |
| "eval_accuracy": 0.38803912068956337, |
| "eval_loss": 3.35626482963562, |
| "eval_runtime": 180.0467, |
| "eval_samples_per_second": 100.035, |
| "eval_steps_per_second": 6.254, |
| "step": 68000 |
| }, |
| { |
| "epoch": 7.324292325906791, |
| "grad_norm": 0.9152792096138, |
| "learning_rate": 0.0001608899903027691, |
| "loss": 3.2376, |
| "step": 68050 |
| }, |
| { |
| "epoch": 7.3296738779464, |
| "grad_norm": 0.770829975605011, |
| "learning_rate": 0.00016056674927270766, |
| "loss": 3.2143, |
| "step": 68100 |
| }, |
| { |
| "epoch": 7.335055429986008, |
| "grad_norm": 0.7539659142494202, |
| "learning_rate": 0.00016024350824264626, |
| "loss": 3.2165, |
| "step": 68150 |
| }, |
| { |
| "epoch": 7.340436982025616, |
| "grad_norm": 0.7711495757102966, |
| "learning_rate": 0.00015992026721258485, |
| "loss": 3.2089, |
| "step": 68200 |
| }, |
| { |
| "epoch": 7.3458185340652244, |
| "grad_norm": 0.8086374402046204, |
| "learning_rate": 0.00015959702618252342, |
| "loss": 3.2291, |
| "step": 68250 |
| }, |
| { |
| "epoch": 7.351200086104833, |
| "grad_norm": 0.7436811327934265, |
| "learning_rate": 0.00015927378515246199, |
| "loss": 3.2341, |
| "step": 68300 |
| }, |
| { |
| "epoch": 7.356581638144441, |
| "grad_norm": 0.7765557169914246, |
| "learning_rate": 0.0001589505441224006, |
| "loss": 3.2137, |
| "step": 68350 |
| }, |
| { |
| "epoch": 7.361963190184049, |
| "grad_norm": 0.8380234837532043, |
| "learning_rate": 0.00015862730309233917, |
| "loss": 3.2272, |
| "step": 68400 |
| }, |
| { |
| "epoch": 7.367344742223658, |
| "grad_norm": 0.7880136966705322, |
| "learning_rate": 0.00015830406206227774, |
| "loss": 3.2346, |
| "step": 68450 |
| }, |
| { |
| "epoch": 7.372726294263265, |
| "grad_norm": 0.7508086562156677, |
| "learning_rate": 0.00015798082103221636, |
| "loss": 3.2303, |
| "step": 68500 |
| }, |
| { |
| "epoch": 7.378107846302874, |
| "grad_norm": 0.7732205390930176, |
| "learning_rate": 0.00015765758000215493, |
| "loss": 3.2266, |
| "step": 68550 |
| }, |
| { |
| "epoch": 7.383489398342482, |
| "grad_norm": 0.7558825016021729, |
| "learning_rate": 0.0001573343389720935, |
| "loss": 3.2174, |
| "step": 68600 |
| }, |
| { |
| "epoch": 7.38887095038209, |
| "grad_norm": 0.8041996359825134, |
| "learning_rate": 0.0001570110979420321, |
| "loss": 3.2201, |
| "step": 68650 |
| }, |
| { |
| "epoch": 7.394252502421699, |
| "grad_norm": 0.7656950354576111, |
| "learning_rate": 0.0001566878569119707, |
| "loss": 3.2176, |
| "step": 68700 |
| }, |
| { |
| "epoch": 7.399634054461306, |
| "grad_norm": 0.7538243532180786, |
| "learning_rate": 0.00015636461588190926, |
| "loss": 3.223, |
| "step": 68750 |
| }, |
| { |
| "epoch": 7.405015606500915, |
| "grad_norm": 0.8169330358505249, |
| "learning_rate": 0.00015604137485184785, |
| "loss": 3.2286, |
| "step": 68800 |
| }, |
| { |
| "epoch": 7.4103971585405235, |
| "grad_norm": 0.7956921458244324, |
| "learning_rate": 0.00015571813382178642, |
| "loss": 3.2345, |
| "step": 68850 |
| }, |
| { |
| "epoch": 7.415778710580131, |
| "grad_norm": 0.7846260070800781, |
| "learning_rate": 0.000155394892791725, |
| "loss": 3.2288, |
| "step": 68900 |
| }, |
| { |
| "epoch": 7.42116026261974, |
| "grad_norm": 0.7715317606925964, |
| "learning_rate": 0.0001550716517616636, |
| "loss": 3.2327, |
| "step": 68950 |
| }, |
| { |
| "epoch": 7.426541814659347, |
| "grad_norm": 0.7787097692489624, |
| "learning_rate": 0.00015474841073160217, |
| "loss": 3.2384, |
| "step": 69000 |
| }, |
| { |
| "epoch": 7.426541814659347, |
| "eval_accuracy": 0.3884835110007246, |
| "eval_loss": 3.3530378341674805, |
| "eval_runtime": 179.8752, |
| "eval_samples_per_second": 100.131, |
| "eval_steps_per_second": 6.26, |
| "step": 69000 |
| }, |
| { |
| "epoch": 7.431923366698956, |
| "grad_norm": 0.8361562490463257, |
| "learning_rate": 0.0001544251697015408, |
| "loss": 3.2302, |
| "step": 69050 |
| }, |
| { |
| "epoch": 7.4373049187385645, |
| "grad_norm": 0.7875263094902039, |
| "learning_rate": 0.00015410192867147936, |
| "loss": 3.2302, |
| "step": 69100 |
| }, |
| { |
| "epoch": 7.442686470778172, |
| "grad_norm": 0.8438063263893127, |
| "learning_rate": 0.00015377868764141793, |
| "loss": 3.2276, |
| "step": 69150 |
| }, |
| { |
| "epoch": 7.448068022817781, |
| "grad_norm": 0.8052544593811035, |
| "learning_rate": 0.00015345544661135653, |
| "loss": 3.2535, |
| "step": 69200 |
| }, |
| { |
| "epoch": 7.453449574857389, |
| "grad_norm": 0.8062599301338196, |
| "learning_rate": 0.00015313220558129512, |
| "loss": 3.2247, |
| "step": 69250 |
| }, |
| { |
| "epoch": 7.458831126896997, |
| "grad_norm": 0.7678873538970947, |
| "learning_rate": 0.0001528089645512337, |
| "loss": 3.2306, |
| "step": 69300 |
| }, |
| { |
| "epoch": 7.4642126789366054, |
| "grad_norm": 0.8218041658401489, |
| "learning_rate": 0.00015248572352117228, |
| "loss": 3.2194, |
| "step": 69350 |
| }, |
| { |
| "epoch": 7.469594230976213, |
| "grad_norm": 0.7693276405334473, |
| "learning_rate": 0.00015216248249111085, |
| "loss": 3.2326, |
| "step": 69400 |
| }, |
| { |
| "epoch": 7.474975783015822, |
| "grad_norm": 0.7829477787017822, |
| "learning_rate": 0.00015183924146104945, |
| "loss": 3.2393, |
| "step": 69450 |
| }, |
| { |
| "epoch": 7.48035733505543, |
| "grad_norm": 0.78624427318573, |
| "learning_rate": 0.00015151600043098804, |
| "loss": 3.2352, |
| "step": 69500 |
| }, |
| { |
| "epoch": 7.485738887095038, |
| "grad_norm": 0.8021953701972961, |
| "learning_rate": 0.0001511927594009266, |
| "loss": 3.23, |
| "step": 69550 |
| }, |
| { |
| "epoch": 7.491120439134646, |
| "grad_norm": 0.8379894495010376, |
| "learning_rate": 0.00015086951837086518, |
| "loss": 3.2369, |
| "step": 69600 |
| }, |
| { |
| "epoch": 7.496501991174255, |
| "grad_norm": 0.8550412654876709, |
| "learning_rate": 0.0001505462773408038, |
| "loss": 3.2208, |
| "step": 69650 |
| }, |
| { |
| "epoch": 7.501883543213863, |
| "grad_norm": 0.7694771885871887, |
| "learning_rate": 0.00015022303631074236, |
| "loss": 3.2236, |
| "step": 69700 |
| }, |
| { |
| "epoch": 7.507265095253471, |
| "grad_norm": 0.8577457666397095, |
| "learning_rate": 0.00014989979528068096, |
| "loss": 3.2487, |
| "step": 69750 |
| }, |
| { |
| "epoch": 7.51264664729308, |
| "grad_norm": 0.7951626181602478, |
| "learning_rate": 0.00014957655425061953, |
| "loss": 3.2151, |
| "step": 69800 |
| }, |
| { |
| "epoch": 7.518028199332687, |
| "grad_norm": 0.770133376121521, |
| "learning_rate": 0.00014925331322055812, |
| "loss": 3.2279, |
| "step": 69850 |
| }, |
| { |
| "epoch": 7.523409751372296, |
| "grad_norm": 0.7697238326072693, |
| "learning_rate": 0.0001489300721904967, |
| "loss": 3.2426, |
| "step": 69900 |
| }, |
| { |
| "epoch": 7.528791303411904, |
| "grad_norm": 0.8447310328483582, |
| "learning_rate": 0.0001486132959810365, |
| "loss": 3.2232, |
| "step": 69950 |
| }, |
| { |
| "epoch": 7.534172855451512, |
| "grad_norm": 0.7980256676673889, |
| "learning_rate": 0.0001482900549509751, |
| "loss": 3.2464, |
| "step": 70000 |
| }, |
| { |
| "epoch": 7.534172855451512, |
| "eval_accuracy": 0.38879534478875216, |
| "eval_loss": 3.348252534866333, |
| "eval_runtime": 180.2429, |
| "eval_samples_per_second": 99.926, |
| "eval_steps_per_second": 6.247, |
| "step": 70000 |
| }, |
| { |
| "epoch": 7.539554407491121, |
| "grad_norm": 0.7334508299827576, |
| "learning_rate": 0.00014796681392091368, |
| "loss": 3.2262, |
| "step": 70050 |
| }, |
| { |
| "epoch": 7.544935959530728, |
| "grad_norm": 0.7756717801094055, |
| "learning_rate": 0.00014764357289085228, |
| "loss": 3.236, |
| "step": 70100 |
| }, |
| { |
| "epoch": 7.550317511570337, |
| "grad_norm": 0.852515459060669, |
| "learning_rate": 0.00014732033186079085, |
| "loss": 3.2377, |
| "step": 70150 |
| }, |
| { |
| "epoch": 7.5556990636099455, |
| "grad_norm": 0.7722263932228088, |
| "learning_rate": 0.00014699709083072944, |
| "loss": 3.222, |
| "step": 70200 |
| }, |
| { |
| "epoch": 7.561080615649553, |
| "grad_norm": 0.8237218856811523, |
| "learning_rate": 0.00014667384980066804, |
| "loss": 3.2207, |
| "step": 70250 |
| }, |
| { |
| "epoch": 7.566462167689162, |
| "grad_norm": 0.7758240699768066, |
| "learning_rate": 0.0001463506087706066, |
| "loss": 3.2207, |
| "step": 70300 |
| }, |
| { |
| "epoch": 7.57184371972877, |
| "grad_norm": 0.7621212601661682, |
| "learning_rate": 0.0001460273677405452, |
| "loss": 3.2295, |
| "step": 70350 |
| }, |
| { |
| "epoch": 7.577225271768378, |
| "grad_norm": 0.7284106016159058, |
| "learning_rate": 0.00014570412671048377, |
| "loss": 3.2325, |
| "step": 70400 |
| }, |
| { |
| "epoch": 7.5826068238079865, |
| "grad_norm": 0.793984055519104, |
| "learning_rate": 0.00014538088568042236, |
| "loss": 3.2337, |
| "step": 70450 |
| }, |
| { |
| "epoch": 7.587988375847594, |
| "grad_norm": 0.7775974869728088, |
| "learning_rate": 0.00014505764465036093, |
| "loss": 3.226, |
| "step": 70500 |
| }, |
| { |
| "epoch": 7.593369927887203, |
| "grad_norm": 0.7938340306282043, |
| "learning_rate": 0.00014473440362029952, |
| "loss": 3.2377, |
| "step": 70550 |
| }, |
| { |
| "epoch": 7.598751479926811, |
| "grad_norm": 0.8512852191925049, |
| "learning_rate": 0.00014441116259023812, |
| "loss": 3.2208, |
| "step": 70600 |
| }, |
| { |
| "epoch": 7.604133031966419, |
| "grad_norm": 0.8293440341949463, |
| "learning_rate": 0.00014408792156017669, |
| "loss": 3.2424, |
| "step": 70650 |
| }, |
| { |
| "epoch": 7.609514584006027, |
| "grad_norm": 0.7990338206291199, |
| "learning_rate": 0.00014376468053011528, |
| "loss": 3.2255, |
| "step": 70700 |
| }, |
| { |
| "epoch": 7.614896136045635, |
| "grad_norm": 0.7645559906959534, |
| "learning_rate": 0.00014344143950005387, |
| "loss": 3.2468, |
| "step": 70750 |
| }, |
| { |
| "epoch": 7.620277688085244, |
| "grad_norm": 0.7902533411979675, |
| "learning_rate": 0.00014311819846999244, |
| "loss": 3.2283, |
| "step": 70800 |
| }, |
| { |
| "epoch": 7.625659240124852, |
| "grad_norm": 0.7919648885726929, |
| "learning_rate": 0.00014279495743993104, |
| "loss": 3.2399, |
| "step": 70850 |
| }, |
| { |
| "epoch": 7.63104079216446, |
| "grad_norm": 0.7276238799095154, |
| "learning_rate": 0.00014247171640986963, |
| "loss": 3.2308, |
| "step": 70900 |
| }, |
| { |
| "epoch": 7.636422344204068, |
| "grad_norm": 0.7745274901390076, |
| "learning_rate": 0.0001421484753798082, |
| "loss": 3.2317, |
| "step": 70950 |
| }, |
| { |
| "epoch": 7.641803896243677, |
| "grad_norm": 0.7933140397071838, |
| "learning_rate": 0.0001418252343497468, |
| "loss": 3.2376, |
| "step": 71000 |
| }, |
| { |
| "epoch": 7.641803896243677, |
| "eval_accuracy": 0.389115218890478, |
| "eval_loss": 3.343194007873535, |
| "eval_runtime": 179.6738, |
| "eval_samples_per_second": 100.243, |
| "eval_steps_per_second": 6.267, |
| "step": 71000 |
| }, |
| { |
| "epoch": 7.647185448283285, |
| "grad_norm": 0.7818636298179626, |
| "learning_rate": 0.00014150199331968536, |
| "loss": 3.2169, |
| "step": 71050 |
| }, |
| { |
| "epoch": 7.652567000322893, |
| "grad_norm": 0.8027185797691345, |
| "learning_rate": 0.00014117875228962396, |
| "loss": 3.219, |
| "step": 71100 |
| }, |
| { |
| "epoch": 7.657948552362502, |
| "grad_norm": 0.8079972863197327, |
| "learning_rate": 0.00014085551125956252, |
| "loss": 3.231, |
| "step": 71150 |
| }, |
| { |
| "epoch": 7.663330104402109, |
| "grad_norm": 0.7901222109794617, |
| "learning_rate": 0.00014053227022950112, |
| "loss": 3.2243, |
| "step": 71200 |
| }, |
| { |
| "epoch": 7.668711656441718, |
| "grad_norm": 0.8354732394218445, |
| "learning_rate": 0.0001402090291994397, |
| "loss": 3.2378, |
| "step": 71250 |
| }, |
| { |
| "epoch": 7.674093208481326, |
| "grad_norm": 0.8050693869590759, |
| "learning_rate": 0.00013988578816937828, |
| "loss": 3.2202, |
| "step": 71300 |
| }, |
| { |
| "epoch": 7.679474760520934, |
| "grad_norm": 0.7609599828720093, |
| "learning_rate": 0.00013956254713931687, |
| "loss": 3.2262, |
| "step": 71350 |
| }, |
| { |
| "epoch": 7.684856312560543, |
| "grad_norm": 0.8210611343383789, |
| "learning_rate": 0.00013923930610925547, |
| "loss": 3.233, |
| "step": 71400 |
| }, |
| { |
| "epoch": 7.69023786460015, |
| "grad_norm": 0.7654331922531128, |
| "learning_rate": 0.00013891606507919404, |
| "loss": 3.2487, |
| "step": 71450 |
| }, |
| { |
| "epoch": 7.695619416639759, |
| "grad_norm": 0.8369812965393066, |
| "learning_rate": 0.00013859282404913263, |
| "loss": 3.2273, |
| "step": 71500 |
| }, |
| { |
| "epoch": 7.7010009686793675, |
| "grad_norm": 0.8057788610458374, |
| "learning_rate": 0.0001382695830190712, |
| "loss": 3.2144, |
| "step": 71550 |
| }, |
| { |
| "epoch": 7.706382520718975, |
| "grad_norm": 0.7989969849586487, |
| "learning_rate": 0.0001379463419890098, |
| "loss": 3.2386, |
| "step": 71600 |
| }, |
| { |
| "epoch": 7.711764072758584, |
| "grad_norm": 0.8108153939247131, |
| "learning_rate": 0.00013762310095894836, |
| "loss": 3.2351, |
| "step": 71650 |
| }, |
| { |
| "epoch": 7.717145624798192, |
| "grad_norm": 0.7994381785392761, |
| "learning_rate": 0.0001373063247494882, |
| "loss": 3.2329, |
| "step": 71700 |
| }, |
| { |
| "epoch": 7.7225271768378, |
| "grad_norm": 0.7698352932929993, |
| "learning_rate": 0.00013698308371942676, |
| "loss": 3.2261, |
| "step": 71750 |
| }, |
| { |
| "epoch": 7.727908728877408, |
| "grad_norm": 0.7884970903396606, |
| "learning_rate": 0.00013665984268936536, |
| "loss": 3.2273, |
| "step": 71800 |
| }, |
| { |
| "epoch": 7.733290280917016, |
| "grad_norm": 0.8238911032676697, |
| "learning_rate": 0.00013633660165930395, |
| "loss": 3.2296, |
| "step": 71850 |
| }, |
| { |
| "epoch": 7.738671832956625, |
| "grad_norm": 0.804165780544281, |
| "learning_rate": 0.00013601336062924255, |
| "loss": 3.2403, |
| "step": 71900 |
| }, |
| { |
| "epoch": 7.744053384996233, |
| "grad_norm": 0.8062986731529236, |
| "learning_rate": 0.00013569011959918111, |
| "loss": 3.2474, |
| "step": 71950 |
| }, |
| { |
| "epoch": 7.749434937035841, |
| "grad_norm": 0.772881805896759, |
| "learning_rate": 0.0001353668785691197, |
| "loss": 3.2447, |
| "step": 72000 |
| }, |
| { |
| "epoch": 7.749434937035841, |
| "eval_accuracy": 0.389392827018844, |
| "eval_loss": 3.343226909637451, |
| "eval_runtime": 180.5115, |
| "eval_samples_per_second": 99.778, |
| "eval_steps_per_second": 6.238, |
| "step": 72000 |
| }, |
| { |
| "epoch": 7.754816489075449, |
| "grad_norm": 0.7907891869544983, |
| "learning_rate": 0.00013504363753905828, |
| "loss": 3.2411, |
| "step": 72050 |
| }, |
| { |
| "epoch": 7.760198041115058, |
| "grad_norm": 0.8008028864860535, |
| "learning_rate": 0.00013472039650899687, |
| "loss": 3.2389, |
| "step": 72100 |
| }, |
| { |
| "epoch": 7.765579593154666, |
| "grad_norm": 0.8094232082366943, |
| "learning_rate": 0.00013439715547893544, |
| "loss": 3.2357, |
| "step": 72150 |
| }, |
| { |
| "epoch": 7.770961145194274, |
| "grad_norm": 0.7857368588447571, |
| "learning_rate": 0.00013407391444887403, |
| "loss": 3.2359, |
| "step": 72200 |
| }, |
| { |
| "epoch": 7.776342697233883, |
| "grad_norm": 0.7756304740905762, |
| "learning_rate": 0.0001337506734188126, |
| "loss": 3.2329, |
| "step": 72250 |
| }, |
| { |
| "epoch": 7.78172424927349, |
| "grad_norm": 0.7991217374801636, |
| "learning_rate": 0.0001334274323887512, |
| "loss": 3.2244, |
| "step": 72300 |
| }, |
| { |
| "epoch": 7.787105801313099, |
| "grad_norm": 0.8120244741439819, |
| "learning_rate": 0.0001331041913586898, |
| "loss": 3.2231, |
| "step": 72350 |
| }, |
| { |
| "epoch": 7.792487353352707, |
| "grad_norm": 0.8280323147773743, |
| "learning_rate": 0.00013278095032862838, |
| "loss": 3.2446, |
| "step": 72400 |
| }, |
| { |
| "epoch": 7.797868905392315, |
| "grad_norm": 0.7918257713317871, |
| "learning_rate": 0.00013245770929856695, |
| "loss": 3.2223, |
| "step": 72450 |
| }, |
| { |
| "epoch": 7.803250457431924, |
| "grad_norm": 0.7558754086494446, |
| "learning_rate": 0.00013213446826850555, |
| "loss": 3.241, |
| "step": 72500 |
| }, |
| { |
| "epoch": 7.808632009471531, |
| "grad_norm": 0.8165222406387329, |
| "learning_rate": 0.00013181122723844411, |
| "loss": 3.2175, |
| "step": 72550 |
| }, |
| { |
| "epoch": 7.81401356151114, |
| "grad_norm": 0.8103949427604675, |
| "learning_rate": 0.0001314879862083827, |
| "loss": 3.2113, |
| "step": 72600 |
| }, |
| { |
| "epoch": 7.819395113550748, |
| "grad_norm": 0.7988500595092773, |
| "learning_rate": 0.00013116474517832128, |
| "loss": 3.2328, |
| "step": 72650 |
| }, |
| { |
| "epoch": 7.824776665590356, |
| "grad_norm": 0.838823139667511, |
| "learning_rate": 0.00013084150414825987, |
| "loss": 3.2343, |
| "step": 72700 |
| }, |
| { |
| "epoch": 7.830158217629965, |
| "grad_norm": 0.8075284957885742, |
| "learning_rate": 0.00013051826311819844, |
| "loss": 3.2446, |
| "step": 72750 |
| }, |
| { |
| "epoch": 7.835539769669572, |
| "grad_norm": 0.7953897714614868, |
| "learning_rate": 0.00013019502208813703, |
| "loss": 3.2419, |
| "step": 72800 |
| }, |
| { |
| "epoch": 7.840921321709181, |
| "grad_norm": 0.7763371467590332, |
| "learning_rate": 0.00012987178105807563, |
| "loss": 3.227, |
| "step": 72850 |
| }, |
| { |
| "epoch": 7.846302873748789, |
| "grad_norm": 0.8023640513420105, |
| "learning_rate": 0.00012954854002801422, |
| "loss": 3.2381, |
| "step": 72900 |
| }, |
| { |
| "epoch": 7.851684425788397, |
| "grad_norm": 0.7901715040206909, |
| "learning_rate": 0.0001292252989979528, |
| "loss": 3.2443, |
| "step": 72950 |
| }, |
| { |
| "epoch": 7.857065977828006, |
| "grad_norm": 0.7858378887176514, |
| "learning_rate": 0.00012890205796789139, |
| "loss": 3.2332, |
| "step": 73000 |
| }, |
| { |
| "epoch": 7.857065977828006, |
| "eval_accuracy": 0.39001747247088947, |
| "eval_loss": 3.336719274520874, |
| "eval_runtime": 179.8309, |
| "eval_samples_per_second": 100.155, |
| "eval_steps_per_second": 6.261, |
| "step": 73000 |
| }, |
| { |
| "epoch": 7.862447529867614, |
| "grad_norm": 0.795928418636322, |
| "learning_rate": 0.00012857881693782998, |
| "loss": 3.2402, |
| "step": 73050 |
| }, |
| { |
| "epoch": 7.867829081907222, |
| "grad_norm": 0.8185012340545654, |
| "learning_rate": 0.00012825557590776855, |
| "loss": 3.2294, |
| "step": 73100 |
| }, |
| { |
| "epoch": 7.87321063394683, |
| "grad_norm": 0.8013267517089844, |
| "learning_rate": 0.00012793233487770714, |
| "loss": 3.2336, |
| "step": 73150 |
| }, |
| { |
| "epoch": 7.878592185986438, |
| "grad_norm": 0.8170375227928162, |
| "learning_rate": 0.0001276090938476457, |
| "loss": 3.2552, |
| "step": 73200 |
| }, |
| { |
| "epoch": 7.883973738026047, |
| "grad_norm": 0.7894675731658936, |
| "learning_rate": 0.0001272858528175843, |
| "loss": 3.2187, |
| "step": 73250 |
| }, |
| { |
| "epoch": 7.889355290065655, |
| "grad_norm": 0.7849202752113342, |
| "learning_rate": 0.00012696261178752287, |
| "loss": 3.2207, |
| "step": 73300 |
| }, |
| { |
| "epoch": 7.894736842105263, |
| "grad_norm": 0.799229621887207, |
| "learning_rate": 0.00012663937075746147, |
| "loss": 3.2559, |
| "step": 73350 |
| }, |
| { |
| "epoch": 7.900118394144871, |
| "grad_norm": 0.7787337303161621, |
| "learning_rate": 0.00012631612972740006, |
| "loss": 3.2399, |
| "step": 73400 |
| }, |
| { |
| "epoch": 7.90549994618448, |
| "grad_norm": 0.7923638224601746, |
| "learning_rate": 0.00012599288869733863, |
| "loss": 3.2334, |
| "step": 73450 |
| }, |
| { |
| "epoch": 7.910881498224088, |
| "grad_norm": 0.7863078713417053, |
| "learning_rate": 0.00012566964766727722, |
| "loss": 3.2534, |
| "step": 73500 |
| }, |
| { |
| "epoch": 7.916263050263696, |
| "grad_norm": 0.8022647500038147, |
| "learning_rate": 0.00012534640663721582, |
| "loss": 3.2506, |
| "step": 73550 |
| }, |
| { |
| "epoch": 7.921644602303305, |
| "grad_norm": 0.8193617463111877, |
| "learning_rate": 0.00012502316560715439, |
| "loss": 3.2194, |
| "step": 73600 |
| }, |
| { |
| "epoch": 7.927026154342912, |
| "grad_norm": 0.7646448612213135, |
| "learning_rate": 0.00012469992457709298, |
| "loss": 3.2243, |
| "step": 73650 |
| }, |
| { |
| "epoch": 7.932407706382521, |
| "grad_norm": 0.7733369469642639, |
| "learning_rate": 0.00012437668354703158, |
| "loss": 3.2191, |
| "step": 73700 |
| }, |
| { |
| "epoch": 7.937789258422129, |
| "grad_norm": 0.8152474164962769, |
| "learning_rate": 0.00012405344251697014, |
| "loss": 3.2421, |
| "step": 73750 |
| }, |
| { |
| "epoch": 7.943170810461737, |
| "grad_norm": 0.786149263381958, |
| "learning_rate": 0.00012373020148690874, |
| "loss": 3.2335, |
| "step": 73800 |
| }, |
| { |
| "epoch": 7.948552362501346, |
| "grad_norm": 0.8432660102844238, |
| "learning_rate": 0.0001234069604568473, |
| "loss": 3.2521, |
| "step": 73850 |
| }, |
| { |
| "epoch": 7.953933914540953, |
| "grad_norm": 0.8004118204116821, |
| "learning_rate": 0.0001230837194267859, |
| "loss": 3.2383, |
| "step": 73900 |
| }, |
| { |
| "epoch": 7.959315466580562, |
| "grad_norm": 0.7988388538360596, |
| "learning_rate": 0.00012276047839672447, |
| "loss": 3.2287, |
| "step": 73950 |
| }, |
| { |
| "epoch": 7.96469701862017, |
| "grad_norm": 0.7863640189170837, |
| "learning_rate": 0.00012243723736666306, |
| "loss": 3.212, |
| "step": 74000 |
| }, |
| { |
| "epoch": 7.96469701862017, |
| "eval_accuracy": 0.39020229103306187, |
| "eval_loss": 3.332491636276245, |
| "eval_runtime": 180.1547, |
| "eval_samples_per_second": 99.975, |
| "eval_steps_per_second": 6.25, |
| "step": 74000 |
| }, |
| { |
| "epoch": 7.970078570659778, |
| "grad_norm": 0.7968112826347351, |
| "learning_rate": 0.00012211399633660166, |
| "loss": 3.2303, |
| "step": 74050 |
| }, |
| { |
| "epoch": 7.975460122699387, |
| "grad_norm": 0.7782838344573975, |
| "learning_rate": 0.00012179075530654022, |
| "loss": 3.2324, |
| "step": 74100 |
| }, |
| { |
| "epoch": 7.980841674738995, |
| "grad_norm": 0.8536900877952576, |
| "learning_rate": 0.00012146751427647882, |
| "loss": 3.2439, |
| "step": 74150 |
| }, |
| { |
| "epoch": 7.986223226778603, |
| "grad_norm": 0.8339260220527649, |
| "learning_rate": 0.0001211442732464174, |
| "loss": 3.2338, |
| "step": 74200 |
| }, |
| { |
| "epoch": 7.991604778818211, |
| "grad_norm": 0.8594132661819458, |
| "learning_rate": 0.00012082103221635598, |
| "loss": 3.2435, |
| "step": 74250 |
| }, |
| { |
| "epoch": 7.996986330857819, |
| "grad_norm": 0.8051967620849609, |
| "learning_rate": 0.00012049779118629456, |
| "loss": 3.2442, |
| "step": 74300 |
| }, |
| { |
| "epoch": 8.002367882897428, |
| "grad_norm": 0.8067830801010132, |
| "learning_rate": 0.00012017455015623316, |
| "loss": 3.1685, |
| "step": 74350 |
| }, |
| { |
| "epoch": 8.007749434937036, |
| "grad_norm": 0.8121262788772583, |
| "learning_rate": 0.00011985130912617175, |
| "loss": 3.1723, |
| "step": 74400 |
| }, |
| { |
| "epoch": 8.013130986976645, |
| "grad_norm": 0.8569990992546082, |
| "learning_rate": 0.00011952806809611032, |
| "loss": 3.1684, |
| "step": 74450 |
| }, |
| { |
| "epoch": 8.018512539016251, |
| "grad_norm": 0.8412384986877441, |
| "learning_rate": 0.00011920482706604891, |
| "loss": 3.1601, |
| "step": 74500 |
| }, |
| { |
| "epoch": 8.02389409105586, |
| "grad_norm": 0.8731394410133362, |
| "learning_rate": 0.0001188815860359875, |
| "loss": 3.1824, |
| "step": 74550 |
| }, |
| { |
| "epoch": 8.029275643095469, |
| "grad_norm": 0.8120042681694031, |
| "learning_rate": 0.00011855834500592608, |
| "loss": 3.1549, |
| "step": 74600 |
| }, |
| { |
| "epoch": 8.034657195135077, |
| "grad_norm": 0.8084120750427246, |
| "learning_rate": 0.00011823510397586466, |
| "loss": 3.1696, |
| "step": 74650 |
| }, |
| { |
| "epoch": 8.040038747174686, |
| "grad_norm": 0.8044456839561462, |
| "learning_rate": 0.00011791186294580325, |
| "loss": 3.1584, |
| "step": 74700 |
| }, |
| { |
| "epoch": 8.045420299214294, |
| "grad_norm": 0.8172317147254944, |
| "learning_rate": 0.00011758862191574182, |
| "loss": 3.1491, |
| "step": 74750 |
| }, |
| { |
| "epoch": 8.050801851253901, |
| "grad_norm": 0.8528711795806885, |
| "learning_rate": 0.00011726538088568041, |
| "loss": 3.1622, |
| "step": 74800 |
| }, |
| { |
| "epoch": 8.05618340329351, |
| "grad_norm": 0.7957918643951416, |
| "learning_rate": 0.000116942139855619, |
| "loss": 3.1509, |
| "step": 74850 |
| }, |
| { |
| "epoch": 8.061564955333118, |
| "grad_norm": 0.8333423137664795, |
| "learning_rate": 0.00011661889882555759, |
| "loss": 3.1685, |
| "step": 74900 |
| }, |
| { |
| "epoch": 8.066946507372727, |
| "grad_norm": 0.8132439851760864, |
| "learning_rate": 0.00011629565779549616, |
| "loss": 3.1513, |
| "step": 74950 |
| }, |
| { |
| "epoch": 8.072328059412335, |
| "grad_norm": 0.7772408127784729, |
| "learning_rate": 0.00011597241676543475, |
| "loss": 3.1428, |
| "step": 75000 |
| }, |
| { |
| "epoch": 8.072328059412335, |
| "eval_accuracy": 0.38983102411539977, |
| "eval_loss": 3.3393983840942383, |
| "eval_runtime": 179.925, |
| "eval_samples_per_second": 100.103, |
| "eval_steps_per_second": 6.258, |
| "step": 75000 |
| }, |
| { |
| "epoch": 8.077709611451942, |
| "grad_norm": 0.8047040700912476, |
| "learning_rate": 0.00011564917573537335, |
| "loss": 3.1564, |
| "step": 75050 |
| }, |
| { |
| "epoch": 8.08309116349155, |
| "grad_norm": 0.8130930662155151, |
| "learning_rate": 0.00011532593470531191, |
| "loss": 3.1592, |
| "step": 75100 |
| }, |
| { |
| "epoch": 8.088472715531159, |
| "grad_norm": 0.7982088923454285, |
| "learning_rate": 0.0001150026936752505, |
| "loss": 3.1624, |
| "step": 75150 |
| }, |
| { |
| "epoch": 8.093854267570768, |
| "grad_norm": 0.8022302389144897, |
| "learning_rate": 0.00011467945264518909, |
| "loss": 3.1626, |
| "step": 75200 |
| }, |
| { |
| "epoch": 8.099235819610376, |
| "grad_norm": 0.7814152240753174, |
| "learning_rate": 0.00011435621161512766, |
| "loss": 3.1724, |
| "step": 75250 |
| }, |
| { |
| "epoch": 8.104617371649983, |
| "grad_norm": 0.8569167852401733, |
| "learning_rate": 0.00011403297058506625, |
| "loss": 3.1842, |
| "step": 75300 |
| }, |
| { |
| "epoch": 8.109998923689592, |
| "grad_norm": 0.8122818470001221, |
| "learning_rate": 0.00011370972955500485, |
| "loss": 3.163, |
| "step": 75350 |
| }, |
| { |
| "epoch": 8.1153804757292, |
| "grad_norm": 0.7728685736656189, |
| "learning_rate": 0.00011338648852494343, |
| "loss": 3.1683, |
| "step": 75400 |
| }, |
| { |
| "epoch": 8.120762027768809, |
| "grad_norm": 0.8322961330413818, |
| "learning_rate": 0.00011306324749488201, |
| "loss": 3.1758, |
| "step": 75450 |
| }, |
| { |
| "epoch": 8.126143579808417, |
| "grad_norm": 0.8351318836212158, |
| "learning_rate": 0.00011274000646482059, |
| "loss": 3.1767, |
| "step": 75500 |
| }, |
| { |
| "epoch": 8.131525131848026, |
| "grad_norm": 0.8136656880378723, |
| "learning_rate": 0.00011241676543475918, |
| "loss": 3.1513, |
| "step": 75550 |
| }, |
| { |
| "epoch": 8.136906683887632, |
| "grad_norm": 0.8244020342826843, |
| "learning_rate": 0.00011209352440469775, |
| "loss": 3.1645, |
| "step": 75600 |
| }, |
| { |
| "epoch": 8.142288235927241, |
| "grad_norm": 0.8306747078895569, |
| "learning_rate": 0.00011177028337463635, |
| "loss": 3.161, |
| "step": 75650 |
| }, |
| { |
| "epoch": 8.14766978796685, |
| "grad_norm": 0.7907711267471313, |
| "learning_rate": 0.00011144704234457493, |
| "loss": 3.1824, |
| "step": 75700 |
| }, |
| { |
| "epoch": 8.153051340006458, |
| "grad_norm": 0.8126032948493958, |
| "learning_rate": 0.00011113026613511473, |
| "loss": 3.161, |
| "step": 75750 |
| }, |
| { |
| "epoch": 8.158432892046067, |
| "grad_norm": 0.8425970077514648, |
| "learning_rate": 0.00011080702510505333, |
| "loss": 3.1704, |
| "step": 75800 |
| }, |
| { |
| "epoch": 8.163814444085673, |
| "grad_norm": 0.8157441020011902, |
| "learning_rate": 0.00011048378407499191, |
| "loss": 3.1555, |
| "step": 75850 |
| }, |
| { |
| "epoch": 8.169195996125282, |
| "grad_norm": 0.8338661789894104, |
| "learning_rate": 0.00011016054304493049, |
| "loss": 3.1604, |
| "step": 75900 |
| }, |
| { |
| "epoch": 8.17457754816489, |
| "grad_norm": 0.797713041305542, |
| "learning_rate": 0.00010983730201486907, |
| "loss": 3.1651, |
| "step": 75950 |
| }, |
| { |
| "epoch": 8.1799591002045, |
| "grad_norm": 0.8311964869499207, |
| "learning_rate": 0.00010951406098480767, |
| "loss": 3.1588, |
| "step": 76000 |
| }, |
| { |
| "epoch": 8.1799591002045, |
| "eval_accuracy": 0.3903796125459751, |
| "eval_loss": 3.3365893363952637, |
| "eval_runtime": 180.0247, |
| "eval_samples_per_second": 100.047, |
| "eval_steps_per_second": 6.255, |
| "step": 76000 |
| }, |
| { |
| "epoch": 8.185340652244108, |
| "grad_norm": 0.8243706226348877, |
| "learning_rate": 0.00010919081995474623, |
| "loss": 3.1717, |
| "step": 76050 |
| }, |
| { |
| "epoch": 8.190722204283716, |
| "grad_norm": 0.8235000371932983, |
| "learning_rate": 0.00010886757892468483, |
| "loss": 3.1711, |
| "step": 76100 |
| }, |
| { |
| "epoch": 8.196103756323323, |
| "grad_norm": 0.8117742538452148, |
| "learning_rate": 0.00010854433789462342, |
| "loss": 3.1864, |
| "step": 76150 |
| }, |
| { |
| "epoch": 8.201485308362932, |
| "grad_norm": 0.820876955986023, |
| "learning_rate": 0.000108221096864562, |
| "loss": 3.1615, |
| "step": 76200 |
| }, |
| { |
| "epoch": 8.20686686040254, |
| "grad_norm": 0.8476978540420532, |
| "learning_rate": 0.00010789785583450059, |
| "loss": 3.1629, |
| "step": 76250 |
| }, |
| { |
| "epoch": 8.212248412442149, |
| "grad_norm": 0.8074275851249695, |
| "learning_rate": 0.00010757461480443917, |
| "loss": 3.1628, |
| "step": 76300 |
| }, |
| { |
| "epoch": 8.217629964481757, |
| "grad_norm": 0.8188003897666931, |
| "learning_rate": 0.00010725137377437776, |
| "loss": 3.1659, |
| "step": 76350 |
| }, |
| { |
| "epoch": 8.223011516521364, |
| "grad_norm": 0.8572345972061157, |
| "learning_rate": 0.00010692813274431633, |
| "loss": 3.1835, |
| "step": 76400 |
| }, |
| { |
| "epoch": 8.228393068560973, |
| "grad_norm": 0.8235215544700623, |
| "learning_rate": 0.00010660489171425492, |
| "loss": 3.1772, |
| "step": 76450 |
| }, |
| { |
| "epoch": 8.233774620600581, |
| "grad_norm": 0.767663836479187, |
| "learning_rate": 0.0001062816506841935, |
| "loss": 3.1676, |
| "step": 76500 |
| }, |
| { |
| "epoch": 8.23915617264019, |
| "grad_norm": 0.8580052852630615, |
| "learning_rate": 0.00010595840965413209, |
| "loss": 3.192, |
| "step": 76550 |
| }, |
| { |
| "epoch": 8.244537724679798, |
| "grad_norm": 0.8200854063034058, |
| "learning_rate": 0.00010563516862407067, |
| "loss": 3.1725, |
| "step": 76600 |
| }, |
| { |
| "epoch": 8.249919276719407, |
| "grad_norm": 0.8221836686134338, |
| "learning_rate": 0.00010531192759400926, |
| "loss": 3.1761, |
| "step": 76650 |
| }, |
| { |
| "epoch": 8.255300828759013, |
| "grad_norm": 0.8396674394607544, |
| "learning_rate": 0.00010498868656394784, |
| "loss": 3.1504, |
| "step": 76700 |
| }, |
| { |
| "epoch": 8.260682380798622, |
| "grad_norm": 0.8755212426185608, |
| "learning_rate": 0.00010466544553388642, |
| "loss": 3.1605, |
| "step": 76750 |
| }, |
| { |
| "epoch": 8.26606393283823, |
| "grad_norm": 0.7833921909332275, |
| "learning_rate": 0.000104342204503825, |
| "loss": 3.1879, |
| "step": 76800 |
| }, |
| { |
| "epoch": 8.27144548487784, |
| "grad_norm": 0.8280619382858276, |
| "learning_rate": 0.0001040189634737636, |
| "loss": 3.1607, |
| "step": 76850 |
| }, |
| { |
| "epoch": 8.276827036917448, |
| "grad_norm": 0.8352559804916382, |
| "learning_rate": 0.00010369572244370217, |
| "loss": 3.1848, |
| "step": 76900 |
| }, |
| { |
| "epoch": 8.282208588957054, |
| "grad_norm": 0.8176780343055725, |
| "learning_rate": 0.00010337248141364076, |
| "loss": 3.1873, |
| "step": 76950 |
| }, |
| { |
| "epoch": 8.287590140996663, |
| "grad_norm": 0.868302583694458, |
| "learning_rate": 0.00010304924038357936, |
| "loss": 3.1633, |
| "step": 77000 |
| }, |
| { |
| "epoch": 8.287590140996663, |
| "eval_accuracy": 0.39057431852093866, |
| "eval_loss": 3.3340141773223877, |
| "eval_runtime": 179.9597, |
| "eval_samples_per_second": 100.084, |
| "eval_steps_per_second": 6.257, |
| "step": 77000 |
| }, |
| { |
| "epoch": 8.292971693036272, |
| "grad_norm": 0.8199362754821777, |
| "learning_rate": 0.00010272599935351792, |
| "loss": 3.1816, |
| "step": 77050 |
| }, |
| { |
| "epoch": 8.29835324507588, |
| "grad_norm": 0.8597607612609863, |
| "learning_rate": 0.00010240275832345652, |
| "loss": 3.182, |
| "step": 77100 |
| }, |
| { |
| "epoch": 8.303734797115489, |
| "grad_norm": 0.8568953275680542, |
| "learning_rate": 0.0001020795172933951, |
| "loss": 3.1675, |
| "step": 77150 |
| }, |
| { |
| "epoch": 8.309116349155097, |
| "grad_norm": 0.8705286979675293, |
| "learning_rate": 0.0001017562762633337, |
| "loss": 3.1871, |
| "step": 77200 |
| }, |
| { |
| "epoch": 8.314497901194704, |
| "grad_norm": 0.8231798410415649, |
| "learning_rate": 0.00010143303523327226, |
| "loss": 3.1839, |
| "step": 77250 |
| }, |
| { |
| "epoch": 8.319879453234313, |
| "grad_norm": 0.8620304465293884, |
| "learning_rate": 0.00010110979420321086, |
| "loss": 3.1756, |
| "step": 77300 |
| }, |
| { |
| "epoch": 8.325261005273921, |
| "grad_norm": 0.8248611688613892, |
| "learning_rate": 0.00010078655317314944, |
| "loss": 3.1995, |
| "step": 77350 |
| }, |
| { |
| "epoch": 8.33064255731353, |
| "grad_norm": 0.8094890117645264, |
| "learning_rate": 0.00010046331214308802, |
| "loss": 3.1779, |
| "step": 77400 |
| }, |
| { |
| "epoch": 8.336024109353138, |
| "grad_norm": 0.844918966293335, |
| "learning_rate": 0.0001001400711130266, |
| "loss": 3.1811, |
| "step": 77450 |
| }, |
| { |
| "epoch": 8.341405661392745, |
| "grad_norm": 0.8428331017494202, |
| "learning_rate": 9.98168300829652e-05, |
| "loss": 3.1706, |
| "step": 77500 |
| }, |
| { |
| "epoch": 8.346787213432354, |
| "grad_norm": 0.8277503252029419, |
| "learning_rate": 9.949358905290376e-05, |
| "loss": 3.1794, |
| "step": 77550 |
| }, |
| { |
| "epoch": 8.352168765471962, |
| "grad_norm": 0.870583713054657, |
| "learning_rate": 9.917034802284236e-05, |
| "loss": 3.1925, |
| "step": 77600 |
| }, |
| { |
| "epoch": 8.35755031751157, |
| "grad_norm": 0.8874440789222717, |
| "learning_rate": 9.884710699278094e-05, |
| "loss": 3.1897, |
| "step": 77650 |
| }, |
| { |
| "epoch": 8.36293186955118, |
| "grad_norm": 0.8165091276168823, |
| "learning_rate": 9.852386596271953e-05, |
| "loss": 3.1606, |
| "step": 77700 |
| }, |
| { |
| "epoch": 8.368313421590786, |
| "grad_norm": 0.8494898080825806, |
| "learning_rate": 9.82006249326581e-05, |
| "loss": 3.189, |
| "step": 77750 |
| }, |
| { |
| "epoch": 8.373694973630395, |
| "grad_norm": 0.8409062027931213, |
| "learning_rate": 9.78773839025967e-05, |
| "loss": 3.1569, |
| "step": 77800 |
| }, |
| { |
| "epoch": 8.379076525670003, |
| "grad_norm": 0.8248782753944397, |
| "learning_rate": 9.755414287253529e-05, |
| "loss": 3.1796, |
| "step": 77850 |
| }, |
| { |
| "epoch": 8.384458077709612, |
| "grad_norm": 0.8239468932151794, |
| "learning_rate": 9.723736666307508e-05, |
| "loss": 3.1884, |
| "step": 77900 |
| }, |
| { |
| "epoch": 8.38983962974922, |
| "grad_norm": 0.8291215896606445, |
| "learning_rate": 9.691412563301368e-05, |
| "loss": 3.1874, |
| "step": 77950 |
| }, |
| { |
| "epoch": 8.395221181788829, |
| "grad_norm": 0.8552483916282654, |
| "learning_rate": 9.659088460295227e-05, |
| "loss": 3.1619, |
| "step": 78000 |
| }, |
| { |
| "epoch": 8.395221181788829, |
| "eval_accuracy": 0.3911120416627325, |
| "eval_loss": 3.330475330352783, |
| "eval_runtime": 179.7478, |
| "eval_samples_per_second": 100.201, |
| "eval_steps_per_second": 6.264, |
| "step": 78000 |
| }, |
| { |
| "epoch": 8.400602733828435, |
| "grad_norm": 0.8288816213607788, |
| "learning_rate": 9.626764357289084e-05, |
| "loss": 3.1793, |
| "step": 78050 |
| }, |
| { |
| "epoch": 8.405984285868044, |
| "grad_norm": 0.8704893589019775, |
| "learning_rate": 9.594440254282943e-05, |
| "loss": 3.1677, |
| "step": 78100 |
| }, |
| { |
| "epoch": 8.411365837907653, |
| "grad_norm": 0.8380488157272339, |
| "learning_rate": 9.562116151276802e-05, |
| "loss": 3.1952, |
| "step": 78150 |
| }, |
| { |
| "epoch": 8.416747389947261, |
| "grad_norm": 0.8037657141685486, |
| "learning_rate": 9.52979204827066e-05, |
| "loss": 3.1903, |
| "step": 78200 |
| }, |
| { |
| "epoch": 8.42212894198687, |
| "grad_norm": 0.8423017263412476, |
| "learning_rate": 9.497467945264518e-05, |
| "loss": 3.1957, |
| "step": 78250 |
| }, |
| { |
| "epoch": 8.427510494026476, |
| "grad_norm": 0.7998765707015991, |
| "learning_rate": 9.465143842258377e-05, |
| "loss": 3.2026, |
| "step": 78300 |
| }, |
| { |
| "epoch": 8.432892046066085, |
| "grad_norm": 0.8594872355461121, |
| "learning_rate": 9.432819739252234e-05, |
| "loss": 3.174, |
| "step": 78350 |
| }, |
| { |
| "epoch": 8.438273598105694, |
| "grad_norm": 0.8684812784194946, |
| "learning_rate": 9.400495636246093e-05, |
| "loss": 3.1623, |
| "step": 78400 |
| }, |
| { |
| "epoch": 8.443655150145302, |
| "grad_norm": 0.8077685236930847, |
| "learning_rate": 9.368171533239952e-05, |
| "loss": 3.1733, |
| "step": 78450 |
| }, |
| { |
| "epoch": 8.44903670218491, |
| "grad_norm": 0.847590982913971, |
| "learning_rate": 9.335847430233811e-05, |
| "loss": 3.1807, |
| "step": 78500 |
| }, |
| { |
| "epoch": 8.45441825422452, |
| "grad_norm": 0.866955041885376, |
| "learning_rate": 9.303523327227668e-05, |
| "loss": 3.1811, |
| "step": 78550 |
| }, |
| { |
| "epoch": 8.459799806264126, |
| "grad_norm": 0.8266087174415588, |
| "learning_rate": 9.271199224221527e-05, |
| "loss": 3.1786, |
| "step": 78600 |
| }, |
| { |
| "epoch": 8.465181358303735, |
| "grad_norm": 0.8498027920722961, |
| "learning_rate": 9.238875121215387e-05, |
| "loss": 3.1924, |
| "step": 78650 |
| }, |
| { |
| "epoch": 8.470562910343343, |
| "grad_norm": 0.9029237627983093, |
| "learning_rate": 9.206551018209243e-05, |
| "loss": 3.192, |
| "step": 78700 |
| }, |
| { |
| "epoch": 8.475944462382952, |
| "grad_norm": 0.8091332912445068, |
| "learning_rate": 9.174226915203103e-05, |
| "loss": 3.1902, |
| "step": 78750 |
| }, |
| { |
| "epoch": 8.48132601442256, |
| "grad_norm": 0.8806012868881226, |
| "learning_rate": 9.141902812196961e-05, |
| "loss": 3.1834, |
| "step": 78800 |
| }, |
| { |
| "epoch": 8.486707566462167, |
| "grad_norm": 0.8003334403038025, |
| "learning_rate": 9.109578709190818e-05, |
| "loss": 3.1967, |
| "step": 78850 |
| }, |
| { |
| "epoch": 8.492089118501776, |
| "grad_norm": 0.8782911896705627, |
| "learning_rate": 9.077254606184677e-05, |
| "loss": 3.1777, |
| "step": 78900 |
| }, |
| { |
| "epoch": 8.497470670541384, |
| "grad_norm": 0.834248960018158, |
| "learning_rate": 9.044930503178537e-05, |
| "loss": 3.1673, |
| "step": 78950 |
| }, |
| { |
| "epoch": 8.502852222580993, |
| "grad_norm": 0.8210733532905579, |
| "learning_rate": 9.012606400172395e-05, |
| "loss": 3.1821, |
| "step": 79000 |
| }, |
| { |
| "epoch": 8.502852222580993, |
| "eval_accuracy": 0.39133499738852784, |
| "eval_loss": 3.3288872241973877, |
| "eval_runtime": 179.8126, |
| "eval_samples_per_second": 100.165, |
| "eval_steps_per_second": 6.262, |
| "step": 79000 |
| }, |
| { |
| "epoch": 8.508233774620601, |
| "grad_norm": 0.8497635126113892, |
| "learning_rate": 8.980282297166253e-05, |
| "loss": 3.1702, |
| "step": 79050 |
| }, |
| { |
| "epoch": 8.513615326660208, |
| "grad_norm": 0.8306612372398376, |
| "learning_rate": 8.947958194160111e-05, |
| "loss": 3.1834, |
| "step": 79100 |
| }, |
| { |
| "epoch": 8.518996878699816, |
| "grad_norm": 0.8895464539527893, |
| "learning_rate": 8.91563409115397e-05, |
| "loss": 3.1958, |
| "step": 79150 |
| }, |
| { |
| "epoch": 8.524378430739425, |
| "grad_norm": 0.8079034686088562, |
| "learning_rate": 8.883309988147827e-05, |
| "loss": 3.1914, |
| "step": 79200 |
| }, |
| { |
| "epoch": 8.529759982779034, |
| "grad_norm": 0.8202755451202393, |
| "learning_rate": 8.850985885141687e-05, |
| "loss": 3.1948, |
| "step": 79250 |
| }, |
| { |
| "epoch": 8.535141534818642, |
| "grad_norm": 0.8289138674736023, |
| "learning_rate": 8.818661782135545e-05, |
| "loss": 3.1863, |
| "step": 79300 |
| }, |
| { |
| "epoch": 8.54052308685825, |
| "grad_norm": 0.8435153961181641, |
| "learning_rate": 8.786337679129403e-05, |
| "loss": 3.1997, |
| "step": 79350 |
| }, |
| { |
| "epoch": 8.545904638897857, |
| "grad_norm": 0.9007829427719116, |
| "learning_rate": 8.754013576123261e-05, |
| "loss": 3.1776, |
| "step": 79400 |
| }, |
| { |
| "epoch": 8.551286190937466, |
| "grad_norm": 0.8751116394996643, |
| "learning_rate": 8.72168947311712e-05, |
| "loss": 3.1819, |
| "step": 79450 |
| }, |
| { |
| "epoch": 8.556667742977075, |
| "grad_norm": 0.8555277585983276, |
| "learning_rate": 8.68936537011098e-05, |
| "loss": 3.1839, |
| "step": 79500 |
| }, |
| { |
| "epoch": 8.562049295016683, |
| "grad_norm": 0.8604739308357239, |
| "learning_rate": 8.657041267104837e-05, |
| "loss": 3.1696, |
| "step": 79550 |
| }, |
| { |
| "epoch": 8.567430847056292, |
| "grad_norm": 0.8286479711532593, |
| "learning_rate": 8.624717164098696e-05, |
| "loss": 3.1646, |
| "step": 79600 |
| }, |
| { |
| "epoch": 8.572812399095898, |
| "grad_norm": 0.826781690120697, |
| "learning_rate": 8.592393061092554e-05, |
| "loss": 3.1811, |
| "step": 79650 |
| }, |
| { |
| "epoch": 8.578193951135507, |
| "grad_norm": 0.8096086978912354, |
| "learning_rate": 8.560068958086412e-05, |
| "loss": 3.1804, |
| "step": 79700 |
| }, |
| { |
| "epoch": 8.583575503175116, |
| "grad_norm": 0.8319725394248962, |
| "learning_rate": 8.528391337140394e-05, |
| "loss": 3.204, |
| "step": 79750 |
| }, |
| { |
| "epoch": 8.588957055214724, |
| "grad_norm": 0.8044587969779968, |
| "learning_rate": 8.496067234134253e-05, |
| "loss": 3.1926, |
| "step": 79800 |
| }, |
| { |
| "epoch": 8.594338607254333, |
| "grad_norm": 0.7995679974555969, |
| "learning_rate": 8.46374313112811e-05, |
| "loss": 3.1776, |
| "step": 79850 |
| }, |
| { |
| "epoch": 8.599720159293941, |
| "grad_norm": 0.8578434586524963, |
| "learning_rate": 8.431419028121969e-05, |
| "loss": 3.1978, |
| "step": 79900 |
| }, |
| { |
| "epoch": 8.605101711333548, |
| "grad_norm": 0.8371725678443909, |
| "learning_rate": 8.399094925115828e-05, |
| "loss": 3.1848, |
| "step": 79950 |
| }, |
| { |
| "epoch": 8.610483263373157, |
| "grad_norm": 0.856256902217865, |
| "learning_rate": 8.366770822109685e-05, |
| "loss": 3.1881, |
| "step": 80000 |
| }, |
| { |
| "epoch": 8.610483263373157, |
| "eval_accuracy": 0.3915523031641568, |
| "eval_loss": 3.3257627487182617, |
| "eval_runtime": 180.2073, |
| "eval_samples_per_second": 99.946, |
| "eval_steps_per_second": 6.248, |
| "step": 80000 |
| }, |
| { |
| "epoch": 8.615864815412765, |
| "grad_norm": 0.8404597640037537, |
| "learning_rate": 8.334446719103544e-05, |
| "loss": 3.1712, |
| "step": 80050 |
| }, |
| { |
| "epoch": 8.621246367452374, |
| "grad_norm": 0.8627851009368896, |
| "learning_rate": 8.302122616097403e-05, |
| "loss": 3.1643, |
| "step": 80100 |
| }, |
| { |
| "epoch": 8.626627919491982, |
| "grad_norm": 0.8004584312438965, |
| "learning_rate": 8.269798513091261e-05, |
| "loss": 3.16, |
| "step": 80150 |
| }, |
| { |
| "epoch": 8.632009471531589, |
| "grad_norm": 0.8076578974723816, |
| "learning_rate": 8.237474410085119e-05, |
| "loss": 3.1925, |
| "step": 80200 |
| }, |
| { |
| "epoch": 8.637391023571197, |
| "grad_norm": 0.8029952645301819, |
| "learning_rate": 8.205150307078978e-05, |
| "loss": 3.2021, |
| "step": 80250 |
| }, |
| { |
| "epoch": 8.642772575610806, |
| "grad_norm": 0.8227396607398987, |
| "learning_rate": 8.172826204072838e-05, |
| "loss": 3.1918, |
| "step": 80300 |
| }, |
| { |
| "epoch": 8.648154127650415, |
| "grad_norm": 0.8576304912567139, |
| "learning_rate": 8.140502101066694e-05, |
| "loss": 3.1858, |
| "step": 80350 |
| }, |
| { |
| "epoch": 8.653535679690023, |
| "grad_norm": 0.802401602268219, |
| "learning_rate": 8.108177998060553e-05, |
| "loss": 3.1806, |
| "step": 80400 |
| }, |
| { |
| "epoch": 8.658917231729632, |
| "grad_norm": 0.9308738112449646, |
| "learning_rate": 8.075853895054412e-05, |
| "loss": 3.1773, |
| "step": 80450 |
| }, |
| { |
| "epoch": 8.664298783769238, |
| "grad_norm": 0.8083286881446838, |
| "learning_rate": 8.043529792048269e-05, |
| "loss": 3.1862, |
| "step": 80500 |
| }, |
| { |
| "epoch": 8.669680335808847, |
| "grad_norm": 0.8586329221725464, |
| "learning_rate": 8.011205689042128e-05, |
| "loss": 3.1954, |
| "step": 80550 |
| }, |
| { |
| "epoch": 8.675061887848456, |
| "grad_norm": 0.8661782741546631, |
| "learning_rate": 7.978881586035988e-05, |
| "loss": 3.1814, |
| "step": 80600 |
| }, |
| { |
| "epoch": 8.680443439888064, |
| "grad_norm": 0.8471584916114807, |
| "learning_rate": 7.946557483029845e-05, |
| "loss": 3.1782, |
| "step": 80650 |
| }, |
| { |
| "epoch": 8.685824991927673, |
| "grad_norm": 0.8519724011421204, |
| "learning_rate": 7.914233380023704e-05, |
| "loss": 3.1871, |
| "step": 80700 |
| }, |
| { |
| "epoch": 8.69120654396728, |
| "grad_norm": 0.883844792842865, |
| "learning_rate": 7.881909277017562e-05, |
| "loss": 3.2014, |
| "step": 80750 |
| }, |
| { |
| "epoch": 8.696588096006888, |
| "grad_norm": 0.8341808915138245, |
| "learning_rate": 7.849585174011422e-05, |
| "loss": 3.1888, |
| "step": 80800 |
| }, |
| { |
| "epoch": 8.701969648046497, |
| "grad_norm": 0.8479406833648682, |
| "learning_rate": 7.817261071005278e-05, |
| "loss": 3.1704, |
| "step": 80850 |
| }, |
| { |
| "epoch": 8.707351200086105, |
| "grad_norm": 0.8445674777030945, |
| "learning_rate": 7.784936967999138e-05, |
| "loss": 3.1814, |
| "step": 80900 |
| }, |
| { |
| "epoch": 8.712732752125714, |
| "grad_norm": 0.876797616481781, |
| "learning_rate": 7.752612864992996e-05, |
| "loss": 3.1678, |
| "step": 80950 |
| }, |
| { |
| "epoch": 8.718114304165322, |
| "grad_norm": 0.8637148141860962, |
| "learning_rate": 7.720288761986854e-05, |
| "loss": 3.1781, |
| "step": 81000 |
| }, |
| { |
| "epoch": 8.718114304165322, |
| "eval_accuracy": 0.3921533709395465, |
| "eval_loss": 3.320990800857544, |
| "eval_runtime": 179.8841, |
| "eval_samples_per_second": 100.126, |
| "eval_steps_per_second": 6.26, |
| "step": 81000 |
| }, |
| { |
| "epoch": 8.723495856204929, |
| "grad_norm": 0.8150270581245422, |
| "learning_rate": 7.687964658980712e-05, |
| "loss": 3.1736, |
| "step": 81050 |
| }, |
| { |
| "epoch": 8.728877408244538, |
| "grad_norm": 0.8342413306236267, |
| "learning_rate": 7.655640555974572e-05, |
| "loss": 3.1732, |
| "step": 81100 |
| }, |
| { |
| "epoch": 8.734258960284146, |
| "grad_norm": 0.8617803454399109, |
| "learning_rate": 7.623316452968428e-05, |
| "loss": 3.1916, |
| "step": 81150 |
| }, |
| { |
| "epoch": 8.739640512323755, |
| "grad_norm": 0.8338009119033813, |
| "learning_rate": 7.590992349962288e-05, |
| "loss": 3.1914, |
| "step": 81200 |
| }, |
| { |
| "epoch": 8.745022064363363, |
| "grad_norm": 0.8587698340415955, |
| "learning_rate": 7.558668246956147e-05, |
| "loss": 3.1983, |
| "step": 81250 |
| }, |
| { |
| "epoch": 8.75040361640297, |
| "grad_norm": 0.8290186524391174, |
| "learning_rate": 7.526344143950005e-05, |
| "loss": 3.1846, |
| "step": 81300 |
| }, |
| { |
| "epoch": 8.755785168442578, |
| "grad_norm": 0.8477612137794495, |
| "learning_rate": 7.494020040943862e-05, |
| "loss": 3.1712, |
| "step": 81350 |
| }, |
| { |
| "epoch": 8.761166720482187, |
| "grad_norm": 0.8442912101745605, |
| "learning_rate": 7.461695937937722e-05, |
| "loss": 3.1715, |
| "step": 81400 |
| }, |
| { |
| "epoch": 8.766548272521796, |
| "grad_norm": 0.8489503264427185, |
| "learning_rate": 7.42937183493158e-05, |
| "loss": 3.1831, |
| "step": 81450 |
| }, |
| { |
| "epoch": 8.771929824561404, |
| "grad_norm": 0.8153248429298401, |
| "learning_rate": 7.397047731925439e-05, |
| "loss": 3.1897, |
| "step": 81500 |
| }, |
| { |
| "epoch": 8.777311376601011, |
| "grad_norm": 0.8930414319038391, |
| "learning_rate": 7.364723628919297e-05, |
| "loss": 3.1778, |
| "step": 81550 |
| }, |
| { |
| "epoch": 8.78269292864062, |
| "grad_norm": 0.845933198928833, |
| "learning_rate": 7.332399525913155e-05, |
| "loss": 3.1975, |
| "step": 81600 |
| }, |
| { |
| "epoch": 8.788074480680228, |
| "grad_norm": 0.8592908382415771, |
| "learning_rate": 7.300075422907013e-05, |
| "loss": 3.1731, |
| "step": 81650 |
| }, |
| { |
| "epoch": 8.793456032719837, |
| "grad_norm": 0.8641341924667358, |
| "learning_rate": 7.267751319900872e-05, |
| "loss": 3.1821, |
| "step": 81700 |
| }, |
| { |
| "epoch": 8.798837584759445, |
| "grad_norm": 0.853470504283905, |
| "learning_rate": 7.235427216894731e-05, |
| "loss": 3.1721, |
| "step": 81750 |
| }, |
| { |
| "epoch": 8.804219136799054, |
| "grad_norm": 0.8515709042549133, |
| "learning_rate": 7.203103113888589e-05, |
| "loss": 3.1734, |
| "step": 81800 |
| }, |
| { |
| "epoch": 8.80960068883866, |
| "grad_norm": 0.802684485912323, |
| "learning_rate": 7.170779010882447e-05, |
| "loss": 3.1746, |
| "step": 81850 |
| }, |
| { |
| "epoch": 8.814982240878269, |
| "grad_norm": 0.8838043212890625, |
| "learning_rate": 7.138454907876305e-05, |
| "loss": 3.1732, |
| "step": 81900 |
| }, |
| { |
| "epoch": 8.820363792917878, |
| "grad_norm": 0.8429666757583618, |
| "learning_rate": 7.106130804870164e-05, |
| "loss": 3.1743, |
| "step": 81950 |
| }, |
| { |
| "epoch": 8.825745344957486, |
| "grad_norm": 0.8476669788360596, |
| "learning_rate": 7.073806701864023e-05, |
| "loss": 3.1834, |
| "step": 82000 |
| }, |
| { |
| "epoch": 8.825745344957486, |
| "eval_accuracy": 0.39253778485663415, |
| "eval_loss": 3.3173086643218994, |
| "eval_runtime": 180.2273, |
| "eval_samples_per_second": 99.935, |
| "eval_steps_per_second": 6.248, |
| "step": 82000 |
| }, |
| { |
| "epoch": 8.831126896997095, |
| "grad_norm": 0.8718939423561096, |
| "learning_rate": 7.041482598857881e-05, |
| "loss": 3.1812, |
| "step": 82050 |
| }, |
| { |
| "epoch": 8.836508449036701, |
| "grad_norm": 0.8387258648872375, |
| "learning_rate": 7.009158495851739e-05, |
| "loss": 3.1846, |
| "step": 82100 |
| }, |
| { |
| "epoch": 8.84189000107631, |
| "grad_norm": 0.8789180517196655, |
| "learning_rate": 6.976834392845599e-05, |
| "loss": 3.178, |
| "step": 82150 |
| }, |
| { |
| "epoch": 8.847271553115919, |
| "grad_norm": 0.8595734238624573, |
| "learning_rate": 6.944510289839457e-05, |
| "loss": 3.1809, |
| "step": 82200 |
| }, |
| { |
| "epoch": 8.852653105155527, |
| "grad_norm": 0.8221034407615662, |
| "learning_rate": 6.912186186833315e-05, |
| "loss": 3.1896, |
| "step": 82250 |
| }, |
| { |
| "epoch": 8.858034657195136, |
| "grad_norm": 0.8374698758125305, |
| "learning_rate": 6.879862083827173e-05, |
| "loss": 3.1773, |
| "step": 82300 |
| }, |
| { |
| "epoch": 8.863416209234742, |
| "grad_norm": 0.8906286358833313, |
| "learning_rate": 6.847537980821031e-05, |
| "loss": 3.1929, |
| "step": 82350 |
| }, |
| { |
| "epoch": 8.868797761274351, |
| "grad_norm": 0.8791503310203552, |
| "learning_rate": 6.81521387781489e-05, |
| "loss": 3.1769, |
| "step": 82400 |
| }, |
| { |
| "epoch": 8.87417931331396, |
| "grad_norm": 0.8439499735832214, |
| "learning_rate": 6.782889774808749e-05, |
| "loss": 3.1919, |
| "step": 82450 |
| }, |
| { |
| "epoch": 8.879560865353568, |
| "grad_norm": 0.8035529851913452, |
| "learning_rate": 6.750565671802607e-05, |
| "loss": 3.1862, |
| "step": 82500 |
| }, |
| { |
| "epoch": 8.884942417393177, |
| "grad_norm": 0.8311242461204529, |
| "learning_rate": 6.718241568796465e-05, |
| "loss": 3.192, |
| "step": 82550 |
| }, |
| { |
| "epoch": 8.890323969432785, |
| "grad_norm": 0.8353226780891418, |
| "learning_rate": 6.685917465790323e-05, |
| "loss": 3.1812, |
| "step": 82600 |
| }, |
| { |
| "epoch": 8.895705521472392, |
| "grad_norm": 0.8998394012451172, |
| "learning_rate": 6.653593362784182e-05, |
| "loss": 3.1869, |
| "step": 82650 |
| }, |
| { |
| "epoch": 8.901087073512, |
| "grad_norm": 0.8866690397262573, |
| "learning_rate": 6.62126925977804e-05, |
| "loss": 3.1857, |
| "step": 82700 |
| }, |
| { |
| "epoch": 8.906468625551609, |
| "grad_norm": 0.9431129097938538, |
| "learning_rate": 6.588945156771899e-05, |
| "loss": 3.1655, |
| "step": 82750 |
| }, |
| { |
| "epoch": 8.911850177591218, |
| "grad_norm": 0.8412008285522461, |
| "learning_rate": 6.556621053765757e-05, |
| "loss": 3.1803, |
| "step": 82800 |
| }, |
| { |
| "epoch": 8.917231729630826, |
| "grad_norm": 0.8523921966552734, |
| "learning_rate": 6.524296950759615e-05, |
| "loss": 3.1745, |
| "step": 82850 |
| }, |
| { |
| "epoch": 8.922613281670433, |
| "grad_norm": 0.8443444967269897, |
| "learning_rate": 6.491972847753474e-05, |
| "loss": 3.2073, |
| "step": 82900 |
| }, |
| { |
| "epoch": 8.927994833710041, |
| "grad_norm": 0.8373326063156128, |
| "learning_rate": 6.459648744747333e-05, |
| "loss": 3.1818, |
| "step": 82950 |
| }, |
| { |
| "epoch": 8.93337638574965, |
| "grad_norm": 0.8930068612098694, |
| "learning_rate": 6.427324641741192e-05, |
| "loss": 3.1769, |
| "step": 83000 |
| }, |
| { |
| "epoch": 8.93337638574965, |
| "eval_accuracy": 0.3927184746090696, |
| "eval_loss": 3.3146286010742188, |
| "eval_runtime": 180.16, |
| "eval_samples_per_second": 99.972, |
| "eval_steps_per_second": 6.25, |
| "step": 83000 |
| }, |
| { |
| "epoch": 8.938757937789259, |
| "grad_norm": 0.850643515586853, |
| "learning_rate": 6.39500053873505e-05, |
| "loss": 3.1891, |
| "step": 83050 |
| }, |
| { |
| "epoch": 8.944139489828867, |
| "grad_norm": 0.8928857445716858, |
| "learning_rate": 6.362676435728908e-05, |
| "loss": 3.1738, |
| "step": 83100 |
| }, |
| { |
| "epoch": 8.949521041868476, |
| "grad_norm": 0.8975667953491211, |
| "learning_rate": 6.330352332722766e-05, |
| "loss": 3.1725, |
| "step": 83150 |
| }, |
| { |
| "epoch": 8.954902593908082, |
| "grad_norm": 0.8075838088989258, |
| "learning_rate": 6.298028229716624e-05, |
| "loss": 3.1837, |
| "step": 83200 |
| }, |
| { |
| "epoch": 8.960284145947691, |
| "grad_norm": 0.8257818222045898, |
| "learning_rate": 6.265704126710484e-05, |
| "loss": 3.1772, |
| "step": 83250 |
| }, |
| { |
| "epoch": 8.9656656979873, |
| "grad_norm": 0.874556303024292, |
| "learning_rate": 6.233380023704342e-05, |
| "loss": 3.188, |
| "step": 83300 |
| }, |
| { |
| "epoch": 8.971047250026908, |
| "grad_norm": 0.9293429255485535, |
| "learning_rate": 6.2010559206982e-05, |
| "loss": 3.1731, |
| "step": 83350 |
| }, |
| { |
| "epoch": 8.976428802066517, |
| "grad_norm": 0.8618453741073608, |
| "learning_rate": 6.168731817692058e-05, |
| "loss": 3.1832, |
| "step": 83400 |
| }, |
| { |
| "epoch": 8.981810354106123, |
| "grad_norm": 0.846990168094635, |
| "learning_rate": 6.136407714685916e-05, |
| "loss": 3.1839, |
| "step": 83450 |
| }, |
| { |
| "epoch": 8.987191906145732, |
| "grad_norm": 0.8998017311096191, |
| "learning_rate": 6.104083611679776e-05, |
| "loss": 3.1788, |
| "step": 83500 |
| }, |
| { |
| "epoch": 8.99257345818534, |
| "grad_norm": 0.8273147940635681, |
| "learning_rate": 6.071759508673634e-05, |
| "loss": 3.1743, |
| "step": 83550 |
| }, |
| { |
| "epoch": 8.997955010224949, |
| "grad_norm": 0.8635324239730835, |
| "learning_rate": 6.039435405667492e-05, |
| "loss": 3.1996, |
| "step": 83600 |
| }, |
| { |
| "epoch": 9.003336562264558, |
| "grad_norm": 0.8398813605308533, |
| "learning_rate": 6.007111302661351e-05, |
| "loss": 3.1242, |
| "step": 83650 |
| }, |
| { |
| "epoch": 9.008718114304166, |
| "grad_norm": 0.858457088470459, |
| "learning_rate": 5.974787199655209e-05, |
| "loss": 3.1269, |
| "step": 83700 |
| }, |
| { |
| "epoch": 9.014099666343773, |
| "grad_norm": 0.8467790484428406, |
| "learning_rate": 5.942463096649068e-05, |
| "loss": 3.1305, |
| "step": 83750 |
| }, |
| { |
| "epoch": 9.019481218383381, |
| "grad_norm": 0.8237690925598145, |
| "learning_rate": 5.910138993642926e-05, |
| "loss": 3.1128, |
| "step": 83800 |
| }, |
| { |
| "epoch": 9.02486277042299, |
| "grad_norm": 0.9006997346878052, |
| "learning_rate": 5.878461372696907e-05, |
| "loss": 3.1345, |
| "step": 83850 |
| }, |
| { |
| "epoch": 9.030244322462599, |
| "grad_norm": 0.899467945098877, |
| "learning_rate": 5.846137269690765e-05, |
| "loss": 3.1302, |
| "step": 83900 |
| }, |
| { |
| "epoch": 9.035625874502207, |
| "grad_norm": 0.8287666440010071, |
| "learning_rate": 5.813813166684624e-05, |
| "loss": 3.1283, |
| "step": 83950 |
| }, |
| { |
| "epoch": 9.041007426541814, |
| "grad_norm": 0.8582010269165039, |
| "learning_rate": 5.781489063678482e-05, |
| "loss": 3.1148, |
| "step": 84000 |
| }, |
| { |
| "epoch": 9.041007426541814, |
| "eval_accuracy": 0.39264361276936544, |
| "eval_loss": 3.318713426589966, |
| "eval_runtime": 179.9421, |
| "eval_samples_per_second": 100.093, |
| "eval_steps_per_second": 6.258, |
| "step": 84000 |
| }, |
| { |
| "epoch": 9.046388978581422, |
| "grad_norm": 0.8633049726486206, |
| "learning_rate": 5.749164960672341e-05, |
| "loss": 3.1233, |
| "step": 84050 |
| }, |
| { |
| "epoch": 9.051770530621031, |
| "grad_norm": 0.8720656037330627, |
| "learning_rate": 5.716840857666199e-05, |
| "loss": 3.1223, |
| "step": 84100 |
| }, |
| { |
| "epoch": 9.05715208266064, |
| "grad_norm": 0.8847928047180176, |
| "learning_rate": 5.684516754660057e-05, |
| "loss": 3.1114, |
| "step": 84150 |
| }, |
| { |
| "epoch": 9.062533634700248, |
| "grad_norm": 0.8395546078681946, |
| "learning_rate": 5.652192651653916e-05, |
| "loss": 3.1071, |
| "step": 84200 |
| }, |
| { |
| "epoch": 9.067915186739857, |
| "grad_norm": 0.8343135118484497, |
| "learning_rate": 5.619868548647774e-05, |
| "loss": 3.1269, |
| "step": 84250 |
| }, |
| { |
| "epoch": 9.073296738779463, |
| "grad_norm": 0.8847300410270691, |
| "learning_rate": 5.5875444456416335e-05, |
| "loss": 3.1341, |
| "step": 84300 |
| }, |
| { |
| "epoch": 9.078678290819072, |
| "grad_norm": 0.857685387134552, |
| "learning_rate": 5.5552203426354916e-05, |
| "loss": 3.1307, |
| "step": 84350 |
| }, |
| { |
| "epoch": 9.08405984285868, |
| "grad_norm": 0.8166672587394714, |
| "learning_rate": 5.52289623962935e-05, |
| "loss": 3.1287, |
| "step": 84400 |
| }, |
| { |
| "epoch": 9.089441394898289, |
| "grad_norm": 0.843408465385437, |
| "learning_rate": 5.4905721366232085e-05, |
| "loss": 3.1302, |
| "step": 84450 |
| }, |
| { |
| "epoch": 9.094822946937898, |
| "grad_norm": 0.8354735970497131, |
| "learning_rate": 5.4582480336170666e-05, |
| "loss": 3.1264, |
| "step": 84500 |
| }, |
| { |
| "epoch": 9.100204498977504, |
| "grad_norm": 0.8715230226516724, |
| "learning_rate": 5.4259239306109254e-05, |
| "loss": 3.1349, |
| "step": 84550 |
| }, |
| { |
| "epoch": 9.105586051017113, |
| "grad_norm": 0.794308066368103, |
| "learning_rate": 5.3935998276047835e-05, |
| "loss": 3.1194, |
| "step": 84600 |
| }, |
| { |
| "epoch": 9.110967603056721, |
| "grad_norm": 0.8556493520736694, |
| "learning_rate": 5.3612757245986416e-05, |
| "loss": 3.1273, |
| "step": 84650 |
| }, |
| { |
| "epoch": 9.11634915509633, |
| "grad_norm": 0.8844385147094727, |
| "learning_rate": 5.3289516215925004e-05, |
| "loss": 3.111, |
| "step": 84700 |
| }, |
| { |
| "epoch": 9.121730707135939, |
| "grad_norm": 0.816949188709259, |
| "learning_rate": 5.2966275185863585e-05, |
| "loss": 3.1177, |
| "step": 84750 |
| }, |
| { |
| "epoch": 9.127112259175545, |
| "grad_norm": 0.8969637751579285, |
| "learning_rate": 5.264303415580217e-05, |
| "loss": 3.1527, |
| "step": 84800 |
| }, |
| { |
| "epoch": 9.132493811215154, |
| "grad_norm": 0.8661097884178162, |
| "learning_rate": 5.2319793125740754e-05, |
| "loss": 3.1297, |
| "step": 84850 |
| }, |
| { |
| "epoch": 9.137875363254762, |
| "grad_norm": 0.8353490233421326, |
| "learning_rate": 5.1996552095679336e-05, |
| "loss": 3.1286, |
| "step": 84900 |
| }, |
| { |
| "epoch": 9.143256915294371, |
| "grad_norm": 0.869312584400177, |
| "learning_rate": 5.1673311065617923e-05, |
| "loss": 3.1292, |
| "step": 84950 |
| }, |
| { |
| "epoch": 9.14863846733398, |
| "grad_norm": 0.8817721605300903, |
| "learning_rate": 5.1350070035556505e-05, |
| "loss": 3.101, |
| "step": 85000 |
| }, |
| { |
| "epoch": 9.14863846733398, |
| "eval_accuracy": 0.39279127204390535, |
| "eval_loss": 3.3183066844940186, |
| "eval_runtime": 180.0381, |
| "eval_samples_per_second": 100.04, |
| "eval_steps_per_second": 6.254, |
| "step": 85000 |
| }, |
| { |
| "epoch": 9.154020019373588, |
| "grad_norm": 0.8361867666244507, |
| "learning_rate": 5.102682900549509e-05, |
| "loss": 3.1389, |
| "step": 85050 |
| }, |
| { |
| "epoch": 9.159401571413195, |
| "grad_norm": 0.8956885933876038, |
| "learning_rate": 5.0703587975433674e-05, |
| "loss": 3.1193, |
| "step": 85100 |
| }, |
| { |
| "epoch": 9.164783123452803, |
| "grad_norm": 0.8367882370948792, |
| "learning_rate": 5.0380346945372255e-05, |
| "loss": 3.1208, |
| "step": 85150 |
| }, |
| { |
| "epoch": 9.170164675492412, |
| "grad_norm": 0.8613442778587341, |
| "learning_rate": 5.005710591531085e-05, |
| "loss": 3.1245, |
| "step": 85200 |
| }, |
| { |
| "epoch": 9.17554622753202, |
| "grad_norm": 0.8361165523529053, |
| "learning_rate": 4.973386488524943e-05, |
| "loss": 3.1413, |
| "step": 85250 |
| }, |
| { |
| "epoch": 9.180927779571629, |
| "grad_norm": 0.8090612292289734, |
| "learning_rate": 4.941062385518802e-05, |
| "loss": 3.1308, |
| "step": 85300 |
| }, |
| { |
| "epoch": 9.186309331611236, |
| "grad_norm": 0.8353068828582764, |
| "learning_rate": 4.90873828251266e-05, |
| "loss": 3.1282, |
| "step": 85350 |
| }, |
| { |
| "epoch": 9.191690883650844, |
| "grad_norm": 0.8467201590538025, |
| "learning_rate": 4.876414179506518e-05, |
| "loss": 3.1291, |
| "step": 85400 |
| }, |
| { |
| "epoch": 9.197072435690453, |
| "grad_norm": 0.9284381866455078, |
| "learning_rate": 4.844090076500377e-05, |
| "loss": 3.138, |
| "step": 85450 |
| }, |
| { |
| "epoch": 9.202453987730062, |
| "grad_norm": 0.8707318902015686, |
| "learning_rate": 4.811765973494235e-05, |
| "loss": 3.1146, |
| "step": 85500 |
| }, |
| { |
| "epoch": 9.20783553976967, |
| "grad_norm": 0.8575342297554016, |
| "learning_rate": 4.779441870488094e-05, |
| "loss": 3.1414, |
| "step": 85550 |
| }, |
| { |
| "epoch": 9.213217091809279, |
| "grad_norm": 0.8592131733894348, |
| "learning_rate": 4.747117767481952e-05, |
| "loss": 3.1118, |
| "step": 85600 |
| }, |
| { |
| "epoch": 9.218598643848885, |
| "grad_norm": 0.868922770023346, |
| "learning_rate": 4.71479366447581e-05, |
| "loss": 3.1254, |
| "step": 85650 |
| }, |
| { |
| "epoch": 9.223980195888494, |
| "grad_norm": 0.8733725547790527, |
| "learning_rate": 4.682469561469669e-05, |
| "loss": 3.1187, |
| "step": 85700 |
| }, |
| { |
| "epoch": 9.229361747928102, |
| "grad_norm": 0.853115439414978, |
| "learning_rate": 4.650145458463527e-05, |
| "loss": 3.1316, |
| "step": 85750 |
| }, |
| { |
| "epoch": 9.234743299967711, |
| "grad_norm": 0.8624739050865173, |
| "learning_rate": 4.6178213554573856e-05, |
| "loss": 3.1467, |
| "step": 85800 |
| }, |
| { |
| "epoch": 9.24012485200732, |
| "grad_norm": 0.830511748790741, |
| "learning_rate": 4.585497252451244e-05, |
| "loss": 3.1283, |
| "step": 85850 |
| }, |
| { |
| "epoch": 9.245506404046926, |
| "grad_norm": 0.843909740447998, |
| "learning_rate": 4.553173149445102e-05, |
| "loss": 3.1194, |
| "step": 85900 |
| }, |
| { |
| "epoch": 9.250887956086535, |
| "grad_norm": 0.8719884753227234, |
| "learning_rate": 4.520849046438961e-05, |
| "loss": 3.1206, |
| "step": 85950 |
| }, |
| { |
| "epoch": 9.256269508126143, |
| "grad_norm": 0.8853352665901184, |
| "learning_rate": 4.488524943432819e-05, |
| "loss": 3.1249, |
| "step": 86000 |
| }, |
| { |
| "epoch": 9.256269508126143, |
| "eval_accuracy": 0.3932848821132465, |
| "eval_loss": 3.315089702606201, |
| "eval_runtime": 179.9252, |
| "eval_samples_per_second": 100.103, |
| "eval_steps_per_second": 6.258, |
| "step": 86000 |
| }, |
| { |
| "epoch": 9.261651060165752, |
| "grad_norm": 0.8370910286903381, |
| "learning_rate": 4.456200840426678e-05, |
| "loss": 3.1212, |
| "step": 86050 |
| }, |
| { |
| "epoch": 9.26703261220536, |
| "grad_norm": 0.8677020072937012, |
| "learning_rate": 4.4245232194806596e-05, |
| "loss": 3.1426, |
| "step": 86100 |
| }, |
| { |
| "epoch": 9.272414164244967, |
| "grad_norm": 0.8646278977394104, |
| "learning_rate": 4.392199116474518e-05, |
| "loss": 3.1316, |
| "step": 86150 |
| }, |
| { |
| "epoch": 9.277795716284576, |
| "grad_norm": 0.8564377427101135, |
| "learning_rate": 4.359875013468376e-05, |
| "loss": 3.1319, |
| "step": 86200 |
| }, |
| { |
| "epoch": 9.283177268324184, |
| "grad_norm": 0.8211710453033447, |
| "learning_rate": 4.3275509104622346e-05, |
| "loss": 3.1192, |
| "step": 86250 |
| }, |
| { |
| "epoch": 9.288558820363793, |
| "grad_norm": 0.8410226702690125, |
| "learning_rate": 4.295226807456093e-05, |
| "loss": 3.1324, |
| "step": 86300 |
| }, |
| { |
| "epoch": 9.293940372403402, |
| "grad_norm": 0.8313528299331665, |
| "learning_rate": 4.2629027044499515e-05, |
| "loss": 3.1334, |
| "step": 86350 |
| }, |
| { |
| "epoch": 9.29932192444301, |
| "grad_norm": 0.8200216889381409, |
| "learning_rate": 4.2305786014438096e-05, |
| "loss": 3.1349, |
| "step": 86400 |
| }, |
| { |
| "epoch": 9.304703476482617, |
| "grad_norm": 0.8536425232887268, |
| "learning_rate": 4.198254498437668e-05, |
| "loss": 3.1515, |
| "step": 86450 |
| }, |
| { |
| "epoch": 9.310085028522225, |
| "grad_norm": 0.8287447690963745, |
| "learning_rate": 4.1659303954315265e-05, |
| "loss": 3.1178, |
| "step": 86500 |
| }, |
| { |
| "epoch": 9.315466580561834, |
| "grad_norm": 0.8226425051689148, |
| "learning_rate": 4.1336062924253846e-05, |
| "loss": 3.1261, |
| "step": 86550 |
| }, |
| { |
| "epoch": 9.320848132601443, |
| "grad_norm": 0.8848347663879395, |
| "learning_rate": 4.1012821894192434e-05, |
| "loss": 3.1234, |
| "step": 86600 |
| }, |
| { |
| "epoch": 9.326229684641051, |
| "grad_norm": 0.8347705006599426, |
| "learning_rate": 4.0689580864131015e-05, |
| "loss": 3.1335, |
| "step": 86650 |
| }, |
| { |
| "epoch": 9.331611236680658, |
| "grad_norm": 0.8725020885467529, |
| "learning_rate": 4.0366339834069596e-05, |
| "loss": 3.1235, |
| "step": 86700 |
| }, |
| { |
| "epoch": 9.336992788720266, |
| "grad_norm": 0.834156334400177, |
| "learning_rate": 4.0043098804008184e-05, |
| "loss": 3.1187, |
| "step": 86750 |
| }, |
| { |
| "epoch": 9.342374340759875, |
| "grad_norm": 0.8515089154243469, |
| "learning_rate": 3.9719857773946765e-05, |
| "loss": 3.1172, |
| "step": 86800 |
| }, |
| { |
| "epoch": 9.347755892799483, |
| "grad_norm": 0.8820292949676514, |
| "learning_rate": 3.939661674388536e-05, |
| "loss": 3.1238, |
| "step": 86850 |
| }, |
| { |
| "epoch": 9.353137444839092, |
| "grad_norm": 0.8537640571594238, |
| "learning_rate": 3.9073375713823934e-05, |
| "loss": 3.1401, |
| "step": 86900 |
| }, |
| { |
| "epoch": 9.3585189968787, |
| "grad_norm": 0.900911808013916, |
| "learning_rate": 3.8750134683762515e-05, |
| "loss": 3.1283, |
| "step": 86950 |
| }, |
| { |
| "epoch": 9.363900548918307, |
| "grad_norm": 0.8385894298553467, |
| "learning_rate": 3.842689365370111e-05, |
| "loss": 3.1282, |
| "step": 87000 |
| }, |
| { |
| "epoch": 9.363900548918307, |
| "eval_accuracy": 0.3934482960565195, |
| "eval_loss": 3.3138954639434814, |
| "eval_runtime": 180.0796, |
| "eval_samples_per_second": 100.017, |
| "eval_steps_per_second": 6.253, |
| "step": 87000 |
| }, |
| { |
| "epoch": 9.369282100957916, |
| "grad_norm": 0.8399181365966797, |
| "learning_rate": 3.810365262363969e-05, |
| "loss": 3.1466, |
| "step": 87050 |
| }, |
| { |
| "epoch": 9.374663652997524, |
| "grad_norm": 0.8344743251800537, |
| "learning_rate": 3.778041159357828e-05, |
| "loss": 3.1362, |
| "step": 87100 |
| }, |
| { |
| "epoch": 9.380045205037133, |
| "grad_norm": 0.8556305170059204, |
| "learning_rate": 3.745717056351686e-05, |
| "loss": 3.1363, |
| "step": 87150 |
| }, |
| { |
| "epoch": 9.385426757076742, |
| "grad_norm": 0.8326317071914673, |
| "learning_rate": 3.713392953345544e-05, |
| "loss": 3.1301, |
| "step": 87200 |
| }, |
| { |
| "epoch": 9.390808309116348, |
| "grad_norm": 0.8645597100257874, |
| "learning_rate": 3.681068850339403e-05, |
| "loss": 3.13, |
| "step": 87250 |
| }, |
| { |
| "epoch": 9.396189861155957, |
| "grad_norm": 0.8261982798576355, |
| "learning_rate": 3.648744747333262e-05, |
| "loss": 3.1417, |
| "step": 87300 |
| }, |
| { |
| "epoch": 9.401571413195565, |
| "grad_norm": 0.8522182106971741, |
| "learning_rate": 3.61642064432712e-05, |
| "loss": 3.1177, |
| "step": 87350 |
| }, |
| { |
| "epoch": 9.406952965235174, |
| "grad_norm": 0.8481505513191223, |
| "learning_rate": 3.584096541320978e-05, |
| "loss": 3.1478, |
| "step": 87400 |
| }, |
| { |
| "epoch": 9.412334517274783, |
| "grad_norm": 0.8410167098045349, |
| "learning_rate": 3.551772438314837e-05, |
| "loss": 3.1244, |
| "step": 87450 |
| }, |
| { |
| "epoch": 9.417716069314391, |
| "grad_norm": 0.8503827452659607, |
| "learning_rate": 3.519448335308695e-05, |
| "loss": 3.1273, |
| "step": 87500 |
| }, |
| { |
| "epoch": 9.423097621353998, |
| "grad_norm": 0.8531398177146912, |
| "learning_rate": 3.4871242323025536e-05, |
| "loss": 3.1317, |
| "step": 87550 |
| }, |
| { |
| "epoch": 9.428479173393606, |
| "grad_norm": 0.8031463027000427, |
| "learning_rate": 3.454800129296412e-05, |
| "loss": 3.1272, |
| "step": 87600 |
| }, |
| { |
| "epoch": 9.433860725433215, |
| "grad_norm": 0.8540205955505371, |
| "learning_rate": 3.42247602629027e-05, |
| "loss": 3.137, |
| "step": 87650 |
| }, |
| { |
| "epoch": 9.439242277472824, |
| "grad_norm": 0.8364588022232056, |
| "learning_rate": 3.3901519232841286e-05, |
| "loss": 3.1339, |
| "step": 87700 |
| }, |
| { |
| "epoch": 9.444623829512432, |
| "grad_norm": 0.8611851930618286, |
| "learning_rate": 3.3578278202779874e-05, |
| "loss": 3.1277, |
| "step": 87750 |
| }, |
| { |
| "epoch": 9.450005381552039, |
| "grad_norm": 0.8790317177772522, |
| "learning_rate": 3.3255037172718455e-05, |
| "loss": 3.1342, |
| "step": 87800 |
| }, |
| { |
| "epoch": 9.455386933591647, |
| "grad_norm": 0.858058750629425, |
| "learning_rate": 3.2931796142657036e-05, |
| "loss": 3.1399, |
| "step": 87850 |
| }, |
| { |
| "epoch": 9.460768485631256, |
| "grad_norm": 0.8198880553245544, |
| "learning_rate": 3.2608555112595624e-05, |
| "loss": 3.1449, |
| "step": 87900 |
| }, |
| { |
| "epoch": 9.466150037670864, |
| "grad_norm": 0.8721171021461487, |
| "learning_rate": 3.2285314082534205e-05, |
| "loss": 3.1154, |
| "step": 87950 |
| }, |
| { |
| "epoch": 9.471531589710473, |
| "grad_norm": 0.8698805570602417, |
| "learning_rate": 3.196207305247279e-05, |
| "loss": 3.1402, |
| "step": 88000 |
| }, |
| { |
| "epoch": 9.471531589710473, |
| "eval_accuracy": 0.39370221785534193, |
| "eval_loss": 3.3107540607452393, |
| "eval_runtime": 180.1755, |
| "eval_samples_per_second": 99.964, |
| "eval_steps_per_second": 6.249, |
| "step": 88000 |
| }, |
| { |
| "epoch": 9.476913141750082, |
| "grad_norm": 0.8367273807525635, |
| "learning_rate": 3.1638832022411374e-05, |
| "loss": 3.14, |
| "step": 88050 |
| }, |
| { |
| "epoch": 9.482294693789688, |
| "grad_norm": 0.8261746764183044, |
| "learning_rate": 3.1315590992349955e-05, |
| "loss": 3.1287, |
| "step": 88100 |
| }, |
| { |
| "epoch": 9.487676245829297, |
| "grad_norm": 0.8505935668945312, |
| "learning_rate": 3.099234996228854e-05, |
| "loss": 3.1379, |
| "step": 88150 |
| }, |
| { |
| "epoch": 9.493057797868905, |
| "grad_norm": 0.8693322539329529, |
| "learning_rate": 3.0675573752828356e-05, |
| "loss": 3.1351, |
| "step": 88200 |
| }, |
| { |
| "epoch": 9.498439349908514, |
| "grad_norm": 0.8701393604278564, |
| "learning_rate": 3.035233272276694e-05, |
| "loss": 3.116, |
| "step": 88250 |
| }, |
| { |
| "epoch": 9.503820901948123, |
| "grad_norm": 0.8780934810638428, |
| "learning_rate": 3.0029091692705525e-05, |
| "loss": 3.138, |
| "step": 88300 |
| }, |
| { |
| "epoch": 9.50920245398773, |
| "grad_norm": 0.8750477433204651, |
| "learning_rate": 2.970585066264411e-05, |
| "loss": 3.113, |
| "step": 88350 |
| }, |
| { |
| "epoch": 9.514584006027338, |
| "grad_norm": 0.8323451280593872, |
| "learning_rate": 2.938260963258269e-05, |
| "loss": 3.1459, |
| "step": 88400 |
| }, |
| { |
| "epoch": 9.519965558066946, |
| "grad_norm": 0.8494437336921692, |
| "learning_rate": 2.9059368602521275e-05, |
| "loss": 3.1316, |
| "step": 88450 |
| }, |
| { |
| "epoch": 9.525347110106555, |
| "grad_norm": 0.8687306046485901, |
| "learning_rate": 2.8736127572459863e-05, |
| "loss": 3.1182, |
| "step": 88500 |
| }, |
| { |
| "epoch": 9.530728662146164, |
| "grad_norm": 0.8358095288276672, |
| "learning_rate": 2.8412886542398448e-05, |
| "loss": 3.1418, |
| "step": 88550 |
| }, |
| { |
| "epoch": 9.536110214185772, |
| "grad_norm": 0.8712921142578125, |
| "learning_rate": 2.8089645512337032e-05, |
| "loss": 3.1184, |
| "step": 88600 |
| }, |
| { |
| "epoch": 9.541491766225379, |
| "grad_norm": 0.9299412965774536, |
| "learning_rate": 2.7766404482275613e-05, |
| "loss": 3.116, |
| "step": 88650 |
| }, |
| { |
| "epoch": 9.546873318264987, |
| "grad_norm": 0.8581258654594421, |
| "learning_rate": 2.7443163452214198e-05, |
| "loss": 3.142, |
| "step": 88700 |
| }, |
| { |
| "epoch": 9.552254870304596, |
| "grad_norm": 0.8373107314109802, |
| "learning_rate": 2.7119922422152782e-05, |
| "loss": 3.108, |
| "step": 88750 |
| }, |
| { |
| "epoch": 9.557636422344205, |
| "grad_norm": 0.83504718542099, |
| "learning_rate": 2.6796681392091367e-05, |
| "loss": 3.1487, |
| "step": 88800 |
| }, |
| { |
| "epoch": 9.563017974383813, |
| "grad_norm": 0.8694041967391968, |
| "learning_rate": 2.6473440362029955e-05, |
| "loss": 3.147, |
| "step": 88850 |
| }, |
| { |
| "epoch": 9.56839952642342, |
| "grad_norm": 0.7938987016677856, |
| "learning_rate": 2.6150199331968536e-05, |
| "loss": 3.1254, |
| "step": 88900 |
| }, |
| { |
| "epoch": 9.573781078463028, |
| "grad_norm": 0.9023017287254333, |
| "learning_rate": 2.582695830190712e-05, |
| "loss": 3.1391, |
| "step": 88950 |
| }, |
| { |
| "epoch": 9.579162630502637, |
| "grad_norm": 0.8764638304710388, |
| "learning_rate": 2.5503717271845705e-05, |
| "loss": 3.1279, |
| "step": 89000 |
| }, |
| { |
| "epoch": 9.579162630502637, |
| "eval_accuracy": 0.39404023698933277, |
| "eval_loss": 3.308558940887451, |
| "eval_runtime": 180.0618, |
| "eval_samples_per_second": 100.027, |
| "eval_steps_per_second": 6.253, |
| "step": 89000 |
| }, |
| { |
| "epoch": 9.584544182542245, |
| "grad_norm": 0.8714621663093567, |
| "learning_rate": 2.518047624178429e-05, |
| "loss": 3.1327, |
| "step": 89050 |
| }, |
| { |
| "epoch": 9.589925734581854, |
| "grad_norm": 0.872740626335144, |
| "learning_rate": 2.4857235211722874e-05, |
| "loss": 3.135, |
| "step": 89100 |
| }, |
| { |
| "epoch": 9.59530728662146, |
| "grad_norm": 0.8976700901985168, |
| "learning_rate": 2.4533994181661455e-05, |
| "loss": 3.1357, |
| "step": 89150 |
| }, |
| { |
| "epoch": 9.60068883866107, |
| "grad_norm": 0.8642706871032715, |
| "learning_rate": 2.421075315160004e-05, |
| "loss": 3.1514, |
| "step": 89200 |
| }, |
| { |
| "epoch": 9.606070390700678, |
| "grad_norm": 0.8876418471336365, |
| "learning_rate": 2.3887512121538624e-05, |
| "loss": 3.1272, |
| "step": 89250 |
| }, |
| { |
| "epoch": 9.611451942740286, |
| "grad_norm": 0.8454718589782715, |
| "learning_rate": 2.3564271091477212e-05, |
| "loss": 3.1263, |
| "step": 89300 |
| }, |
| { |
| "epoch": 9.616833494779895, |
| "grad_norm": 0.8721873760223389, |
| "learning_rate": 2.3241030061415796e-05, |
| "loss": 3.133, |
| "step": 89350 |
| }, |
| { |
| "epoch": 9.622215046819504, |
| "grad_norm": 0.8209410905838013, |
| "learning_rate": 2.2917789031354377e-05, |
| "loss": 3.1391, |
| "step": 89400 |
| }, |
| { |
| "epoch": 9.62759659885911, |
| "grad_norm": 0.8578438758850098, |
| "learning_rate": 2.2594548001292962e-05, |
| "loss": 3.1284, |
| "step": 89450 |
| }, |
| { |
| "epoch": 9.632978150898719, |
| "grad_norm": 0.8609829545021057, |
| "learning_rate": 2.2271306971231546e-05, |
| "loss": 3.1445, |
| "step": 89500 |
| }, |
| { |
| "epoch": 9.638359702938327, |
| "grad_norm": 0.8142579793930054, |
| "learning_rate": 2.194806594117013e-05, |
| "loss": 3.1321, |
| "step": 89550 |
| }, |
| { |
| "epoch": 9.643741254977936, |
| "grad_norm": 0.9409341812133789, |
| "learning_rate": 2.1624824911108715e-05, |
| "loss": 3.1258, |
| "step": 89600 |
| }, |
| { |
| "epoch": 9.649122807017545, |
| "grad_norm": 0.8790959119796753, |
| "learning_rate": 2.1301583881047296e-05, |
| "loss": 3.1316, |
| "step": 89650 |
| }, |
| { |
| "epoch": 9.654504359057151, |
| "grad_norm": 0.8953224420547485, |
| "learning_rate": 2.097834285098588e-05, |
| "loss": 3.1157, |
| "step": 89700 |
| }, |
| { |
| "epoch": 9.65988591109676, |
| "grad_norm": 0.8896967172622681, |
| "learning_rate": 2.065510182092447e-05, |
| "loss": 3.1224, |
| "step": 89750 |
| }, |
| { |
| "epoch": 9.665267463136368, |
| "grad_norm": 0.8611626625061035, |
| "learning_rate": 2.0331860790863053e-05, |
| "loss": 3.1174, |
| "step": 89800 |
| }, |
| { |
| "epoch": 9.670649015175977, |
| "grad_norm": 0.8367219567298889, |
| "learning_rate": 2.0008619760801638e-05, |
| "loss": 3.1503, |
| "step": 89850 |
| }, |
| { |
| "epoch": 9.676030567215586, |
| "grad_norm": 0.8824328184127808, |
| "learning_rate": 1.968537873074022e-05, |
| "loss": 3.1359, |
| "step": 89900 |
| }, |
| { |
| "epoch": 9.681412119255192, |
| "grad_norm": 0.8521384000778198, |
| "learning_rate": 1.9362137700678803e-05, |
| "loss": 3.1371, |
| "step": 89950 |
| }, |
| { |
| "epoch": 9.6867936712948, |
| "grad_norm": 0.8393240571022034, |
| "learning_rate": 1.9038896670617388e-05, |
| "loss": 3.159, |
| "step": 90000 |
| }, |
| { |
| "epoch": 9.6867936712948, |
| "eval_accuracy": 0.3942384198567064, |
| "eval_loss": 3.3065290451049805, |
| "eval_runtime": 180.214, |
| "eval_samples_per_second": 99.942, |
| "eval_steps_per_second": 6.248, |
| "step": 90000 |
| }, |
| { |
| "epoch": 9.69217522333441, |
| "grad_norm": 0.8758241534233093, |
| "learning_rate": 1.8715655640555972e-05, |
| "loss": 3.1504, |
| "step": 90050 |
| }, |
| { |
| "epoch": 9.697556775374018, |
| "grad_norm": 0.8368741273880005, |
| "learning_rate": 1.8392414610494557e-05, |
| "loss": 3.1238, |
| "step": 90100 |
| }, |
| { |
| "epoch": 9.702938327413626, |
| "grad_norm": 0.8934403657913208, |
| "learning_rate": 1.806917358043314e-05, |
| "loss": 3.1538, |
| "step": 90150 |
| }, |
| { |
| "epoch": 9.708319879453235, |
| "grad_norm": 0.9281708598136902, |
| "learning_rate": 1.7745932550371726e-05, |
| "loss": 3.1454, |
| "step": 90200 |
| }, |
| { |
| "epoch": 9.713701431492842, |
| "grad_norm": 0.8880597949028015, |
| "learning_rate": 1.742269152031031e-05, |
| "loss": 3.128, |
| "step": 90250 |
| }, |
| { |
| "epoch": 9.71908298353245, |
| "grad_norm": 0.8629451394081116, |
| "learning_rate": 1.7105915310850124e-05, |
| "loss": 3.1369, |
| "step": 90300 |
| }, |
| { |
| "epoch": 9.724464535572059, |
| "grad_norm": 0.8905329704284668, |
| "learning_rate": 1.6782674280788708e-05, |
| "loss": 3.1219, |
| "step": 90350 |
| }, |
| { |
| "epoch": 9.729846087611667, |
| "grad_norm": 0.9040992856025696, |
| "learning_rate": 1.645943325072729e-05, |
| "loss": 3.1318, |
| "step": 90400 |
| }, |
| { |
| "epoch": 9.735227639651276, |
| "grad_norm": 0.8780677318572998, |
| "learning_rate": 1.6136192220665877e-05, |
| "loss": 3.1296, |
| "step": 90450 |
| }, |
| { |
| "epoch": 9.740609191690883, |
| "grad_norm": 0.8315309882164001, |
| "learning_rate": 1.5812951190604458e-05, |
| "loss": 3.1468, |
| "step": 90500 |
| }, |
| { |
| "epoch": 9.745990743730491, |
| "grad_norm": 0.8430354595184326, |
| "learning_rate": 1.5489710160543043e-05, |
| "loss": 3.1279, |
| "step": 90550 |
| }, |
| { |
| "epoch": 9.7513722957701, |
| "grad_norm": 0.8890626430511475, |
| "learning_rate": 1.5166469130481629e-05, |
| "loss": 3.1419, |
| "step": 90600 |
| }, |
| { |
| "epoch": 9.756753847809708, |
| "grad_norm": 0.8835778832435608, |
| "learning_rate": 1.4843228100420212e-05, |
| "loss": 3.1615, |
| "step": 90650 |
| }, |
| { |
| "epoch": 9.762135399849317, |
| "grad_norm": 0.8142911791801453, |
| "learning_rate": 1.4519987070358796e-05, |
| "loss": 3.1359, |
| "step": 90700 |
| }, |
| { |
| "epoch": 9.767516951888926, |
| "grad_norm": 0.8582029938697815, |
| "learning_rate": 1.419674604029738e-05, |
| "loss": 3.1239, |
| "step": 90750 |
| }, |
| { |
| "epoch": 9.772898503928532, |
| "grad_norm": 0.9165640473365784, |
| "learning_rate": 1.3873505010235965e-05, |
| "loss": 3.1209, |
| "step": 90800 |
| }, |
| { |
| "epoch": 9.77828005596814, |
| "grad_norm": 0.8222701549530029, |
| "learning_rate": 1.355026398017455e-05, |
| "loss": 3.1337, |
| "step": 90850 |
| }, |
| { |
| "epoch": 9.78366160800775, |
| "grad_norm": 0.8537663221359253, |
| "learning_rate": 1.3227022950113132e-05, |
| "loss": 3.14, |
| "step": 90900 |
| }, |
| { |
| "epoch": 9.789043160047358, |
| "grad_norm": 0.8751277923583984, |
| "learning_rate": 1.2903781920051719e-05, |
| "loss": 3.1155, |
| "step": 90950 |
| }, |
| { |
| "epoch": 9.794424712086967, |
| "grad_norm": 0.8880753517150879, |
| "learning_rate": 1.2580540889990301e-05, |
| "loss": 3.1357, |
| "step": 91000 |
| }, |
| { |
| "epoch": 9.794424712086967, |
| "eval_accuracy": 0.3946130550138907, |
| "eval_loss": 3.3034253120422363, |
| "eval_runtime": 180.0837, |
| "eval_samples_per_second": 100.015, |
| "eval_steps_per_second": 6.253, |
| "step": 91000 |
| }, |
| { |
| "epoch": 9.799806264126573, |
| "grad_norm": 0.8514898419380188, |
| "learning_rate": 1.2257299859928886e-05, |
| "loss": 3.1331, |
| "step": 91050 |
| }, |
| { |
| "epoch": 9.805187816166182, |
| "grad_norm": 0.8708495497703552, |
| "learning_rate": 1.193405882986747e-05, |
| "loss": 3.1113, |
| "step": 91100 |
| }, |
| { |
| "epoch": 9.81056936820579, |
| "grad_norm": 0.8504621982574463, |
| "learning_rate": 1.1610817799806053e-05, |
| "loss": 3.1098, |
| "step": 91150 |
| }, |
| { |
| "epoch": 9.815950920245399, |
| "grad_norm": 0.8342603445053101, |
| "learning_rate": 1.128757676974464e-05, |
| "loss": 3.1181, |
| "step": 91200 |
| }, |
| { |
| "epoch": 9.821332472285007, |
| "grad_norm": 0.8990119099617004, |
| "learning_rate": 1.0964335739683222e-05, |
| "loss": 3.1207, |
| "step": 91250 |
| }, |
| { |
| "epoch": 9.826714024324616, |
| "grad_norm": 0.8407130241394043, |
| "learning_rate": 1.0641094709621807e-05, |
| "loss": 3.124, |
| "step": 91300 |
| }, |
| { |
| "epoch": 9.832095576364223, |
| "grad_norm": 0.8398468494415283, |
| "learning_rate": 1.0317853679560393e-05, |
| "loss": 3.1244, |
| "step": 91350 |
| }, |
| { |
| "epoch": 9.837477128403831, |
| "grad_norm": 0.8928409814834595, |
| "learning_rate": 9.994612649498976e-06, |
| "loss": 3.1233, |
| "step": 91400 |
| }, |
| { |
| "epoch": 9.84285868044344, |
| "grad_norm": 0.8549565672874451, |
| "learning_rate": 9.67137161943756e-06, |
| "loss": 3.1248, |
| "step": 91450 |
| }, |
| { |
| "epoch": 9.848240232483048, |
| "grad_norm": 0.8319708108901978, |
| "learning_rate": 9.348130589376145e-06, |
| "loss": 3.1021, |
| "step": 91500 |
| }, |
| { |
| "epoch": 9.853621784522657, |
| "grad_norm": 0.8330086469650269, |
| "learning_rate": 9.024889559314727e-06, |
| "loss": 3.1326, |
| "step": 91550 |
| }, |
| { |
| "epoch": 9.859003336562264, |
| "grad_norm": 0.8426081538200378, |
| "learning_rate": 8.701648529253312e-06, |
| "loss": 3.1118, |
| "step": 91600 |
| }, |
| { |
| "epoch": 9.864384888601872, |
| "grad_norm": 0.8548322916030884, |
| "learning_rate": 8.378407499191896e-06, |
| "loss": 3.1311, |
| "step": 91650 |
| }, |
| { |
| "epoch": 9.869766440641481, |
| "grad_norm": 0.8335317969322205, |
| "learning_rate": 8.055166469130481e-06, |
| "loss": 3.1387, |
| "step": 91700 |
| }, |
| { |
| "epoch": 9.87514799268109, |
| "grad_norm": 0.9084175229072571, |
| "learning_rate": 7.731925439069065e-06, |
| "loss": 3.1367, |
| "step": 91750 |
| }, |
| { |
| "epoch": 9.880529544720698, |
| "grad_norm": 0.8292434215545654, |
| "learning_rate": 7.408684409007649e-06, |
| "loss": 3.1127, |
| "step": 91800 |
| }, |
| { |
| "epoch": 9.885911096760307, |
| "grad_norm": 0.8056044578552246, |
| "learning_rate": 7.085443378946234e-06, |
| "loss": 3.1378, |
| "step": 91850 |
| }, |
| { |
| "epoch": 9.891292648799913, |
| "grad_norm": 0.8410173654556274, |
| "learning_rate": 6.762202348884817e-06, |
| "loss": 3.1161, |
| "step": 91900 |
| }, |
| { |
| "epoch": 9.896674200839522, |
| "grad_norm": 0.9032399654388428, |
| "learning_rate": 6.438961318823403e-06, |
| "loss": 3.1386, |
| "step": 91950 |
| }, |
| { |
| "epoch": 9.90205575287913, |
| "grad_norm": 0.8298187851905823, |
| "learning_rate": 6.115720288761986e-06, |
| "loss": 3.1212, |
| "step": 92000 |
| }, |
| { |
| "epoch": 9.90205575287913, |
| "eval_accuracy": 0.39463576346744395, |
| "eval_loss": 3.303168535232544, |
| "eval_runtime": 180.0813, |
| "eval_samples_per_second": 100.016, |
| "eval_steps_per_second": 6.253, |
| "step": 92000 |
| }, |
| { |
| "epoch": 9.907437304918739, |
| "grad_norm": 0.8719784021377563, |
| "learning_rate": 5.792479258700571e-06, |
| "loss": 3.1241, |
| "step": 92050 |
| }, |
| { |
| "epoch": 9.912818856958348, |
| "grad_norm": 0.8631389141082764, |
| "learning_rate": 5.469238228639154e-06, |
| "loss": 3.1118, |
| "step": 92100 |
| }, |
| { |
| "epoch": 9.918200408997954, |
| "grad_norm": 0.8735671639442444, |
| "learning_rate": 5.145997198577739e-06, |
| "loss": 3.1353, |
| "step": 92150 |
| }, |
| { |
| "epoch": 9.923581961037563, |
| "grad_norm": 0.8469117879867554, |
| "learning_rate": 4.822756168516323e-06, |
| "loss": 3.128, |
| "step": 92200 |
| }, |
| { |
| "epoch": 9.928963513077171, |
| "grad_norm": 0.8447126150131226, |
| "learning_rate": 4.499515138454908e-06, |
| "loss": 3.1262, |
| "step": 92250 |
| }, |
| { |
| "epoch": 9.93434506511678, |
| "grad_norm": 0.8380026817321777, |
| "learning_rate": 4.1762741083934915e-06, |
| "loss": 3.1138, |
| "step": 92300 |
| }, |
| { |
| "epoch": 9.939726617156388, |
| "grad_norm": 0.838094174861908, |
| "learning_rate": 3.853033078332076e-06, |
| "loss": 3.1297, |
| "step": 92350 |
| }, |
| { |
| "epoch": 9.945108169195997, |
| "grad_norm": 0.9846398830413818, |
| "learning_rate": 3.52979204827066e-06, |
| "loss": 3.1255, |
| "step": 92400 |
| }, |
| { |
| "epoch": 9.950489721235604, |
| "grad_norm": 0.8288633823394775, |
| "learning_rate": 3.206551018209244e-06, |
| "loss": 3.1255, |
| "step": 92450 |
| }, |
| { |
| "epoch": 9.955871273275212, |
| "grad_norm": 0.8590484261512756, |
| "learning_rate": 2.8833099881478286e-06, |
| "loss": 3.1037, |
| "step": 92500 |
| }, |
| { |
| "epoch": 9.961252825314821, |
| "grad_norm": 0.8735798001289368, |
| "learning_rate": 2.5600689580864127e-06, |
| "loss": 3.1176, |
| "step": 92550 |
| }, |
| { |
| "epoch": 9.96663437735443, |
| "grad_norm": 0.8541606068611145, |
| "learning_rate": 2.2368279280249972e-06, |
| "loss": 3.1224, |
| "step": 92600 |
| }, |
| { |
| "epoch": 9.972015929394038, |
| "grad_norm": 0.8478603363037109, |
| "learning_rate": 1.9135868979635813e-06, |
| "loss": 3.1271, |
| "step": 92650 |
| }, |
| { |
| "epoch": 9.977397481433645, |
| "grad_norm": 0.8639833331108093, |
| "learning_rate": 1.596810688503394e-06, |
| "loss": 3.1237, |
| "step": 92700 |
| }, |
| { |
| "epoch": 9.982779033473253, |
| "grad_norm": 0.8723976612091064, |
| "learning_rate": 1.2735696584419782e-06, |
| "loss": 3.1151, |
| "step": 92750 |
| }, |
| { |
| "epoch": 9.988160585512862, |
| "grad_norm": 0.8547508716583252, |
| "learning_rate": 9.503286283805623e-07, |
| "loss": 3.1201, |
| "step": 92800 |
| }, |
| { |
| "epoch": 9.99354213755247, |
| "grad_norm": 0.8455697894096375, |
| "learning_rate": 6.270875983191466e-07, |
| "loss": 3.1272, |
| "step": 92850 |
| }, |
| { |
| "epoch": 9.998923689592079, |
| "grad_norm": 0.8672316670417786, |
| "learning_rate": 3.038465682577308e-07, |
| "loss": 3.1172, |
| "step": 92900 |
| }, |
| { |
| "epoch": 10.0, |
| "step": 92910, |
| "total_flos": 7.7683166281728e+17, |
| "train_loss": 3.4544442354291984, |
| "train_runtime": 80531.2369, |
| "train_samples_per_second": 36.918, |
| "train_steps_per_second": 1.154 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 92910, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "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": 7.7683166281728e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
