| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 6.0, |
| "eval_steps": 500, |
| "global_step": 3426, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.035056967572304996, |
| "grad_norm": 12.451324462890625, |
| "learning_rate": 1.8446601941747574e-06, |
| "loss": 3.237, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.07011393514460999, |
| "grad_norm": 4.131791114807129, |
| "learning_rate": 3.7864077669902915e-06, |
| "loss": 2.5266, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.10517090271691498, |
| "grad_norm": 2.2451958656311035, |
| "learning_rate": 5.728155339805825e-06, |
| "loss": 1.8755, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.14022787028921999, |
| "grad_norm": 1.4421552419662476, |
| "learning_rate": 7.66990291262136e-06, |
| "loss": 1.3649, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.175284837861525, |
| "grad_norm": 0.9530179500579834, |
| "learning_rate": 9.611650485436894e-06, |
| "loss": 1.0674, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.21034180543382996, |
| "grad_norm": 0.7194477319717407, |
| "learning_rate": 9.99942798060303e-06, |
| "loss": 0.9241, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.24539877300613497, |
| "grad_norm": 0.6556061506271362, |
| "learning_rate": 9.997104376116195e-06, |
| "loss": 0.8575, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.28045574057843997, |
| "grad_norm": 0.5718048810958862, |
| "learning_rate": 9.992994265395959e-06, |
| "loss": 0.829, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.31551270815074495, |
| "grad_norm": 0.4922148287296295, |
| "learning_rate": 9.987099117840969e-06, |
| "loss": 0.8034, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.35056967572305, |
| "grad_norm": 0.47302234172821045, |
| "learning_rate": 9.979421041015336e-06, |
| "loss": 0.7839, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.38562664329535495, |
| "grad_norm": 0.49009189009666443, |
| "learning_rate": 9.969962779895172e-06, |
| "loss": 0.768, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.42068361086765993, |
| "grad_norm": 0.4963654577732086, |
| "learning_rate": 9.958727715887218e-06, |
| "loss": 0.7628, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.45574057843996496, |
| "grad_norm": 0.5206854343414307, |
| "learning_rate": 9.94571986561998e-06, |
| "loss": 0.7488, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.49079754601226994, |
| "grad_norm": 0.48924869298934937, |
| "learning_rate": 9.930943879507748e-06, |
| "loss": 0.7436, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.5258545135845749, |
| "grad_norm": 0.43540337681770325, |
| "learning_rate": 9.914405040088026e-06, |
| "loss": 0.7375, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5609114811568799, |
| "grad_norm": 0.44258421659469604, |
| "learning_rate": 9.896109260132993e-06, |
| "loss": 0.7277, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.595968448729185, |
| "grad_norm": 0.4955386519432068, |
| "learning_rate": 9.876063080535627e-06, |
| "loss": 0.7284, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.6310254163014899, |
| "grad_norm": 0.5027541518211365, |
| "learning_rate": 9.85427366797129e-06, |
| "loss": 0.7231, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.6660823838737949, |
| "grad_norm": 0.4675957262516022, |
| "learning_rate": 9.830748812335576e-06, |
| "loss": 0.7212, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.7011393514461, |
| "grad_norm": 0.4283595383167267, |
| "learning_rate": 9.805496923959363e-06, |
| "loss": 0.7164, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7361963190184049, |
| "grad_norm": 0.452084481716156, |
| "learning_rate": 9.778527030602049e-06, |
| "loss": 0.711, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.7712532865907099, |
| "grad_norm": 0.4737929105758667, |
| "learning_rate": 9.74984877422405e-06, |
| "loss": 0.7084, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.8063102541630149, |
| "grad_norm": 0.4964485466480255, |
| "learning_rate": 9.719472407539725e-06, |
| "loss": 0.7028, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.8413672217353199, |
| "grad_norm": 0.44363030791282654, |
| "learning_rate": 9.68740879035194e-06, |
| "loss": 0.7045, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.8764241893076249, |
| "grad_norm": 0.5004998445510864, |
| "learning_rate": 9.6536693856696e-06, |
| "loss": 0.6937, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.9114811568799299, |
| "grad_norm": 0.4564264118671417, |
| "learning_rate": 9.618266255609533e-06, |
| "loss": 0.699, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.9465381244522348, |
| "grad_norm": 0.4558616280555725, |
| "learning_rate": 9.58121205708418e-06, |
| "loss": 0.691, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.9815950920245399, |
| "grad_norm": 0.413114458322525, |
| "learning_rate": 9.542520037276636e-06, |
| "loss": 0.6891, |
| "step": 560 |
| }, |
| { |
| "epoch": 1.0157756354075373, |
| "grad_norm": 0.403679758310318, |
| "learning_rate": 9.502204028904687e-06, |
| "loss": 0.6812, |
| "step": 580 |
| }, |
| { |
| "epoch": 1.0508326029798423, |
| "grad_norm": 0.40308722853660583, |
| "learning_rate": 9.46027844527549e-06, |
| "loss": 0.6791, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.0858895705521472, |
| "grad_norm": 0.4085083603858948, |
| "learning_rate": 9.416758275132693e-06, |
| "loss": 0.6803, |
| "step": 620 |
| }, |
| { |
| "epoch": 1.1209465381244523, |
| "grad_norm": 0.4475920796394348, |
| "learning_rate": 9.371659077297843e-06, |
| "loss": 0.6789, |
| "step": 640 |
| }, |
| { |
| "epoch": 1.1560035056967572, |
| "grad_norm": 0.4604188799858093, |
| "learning_rate": 9.324996975107978e-06, |
| "loss": 0.674, |
| "step": 660 |
| }, |
| { |
| "epoch": 1.1910604732690622, |
| "grad_norm": 0.4190482795238495, |
| "learning_rate": 9.276788650651392e-06, |
| "loss": 0.6746, |
| "step": 680 |
| }, |
| { |
| "epoch": 1.2261174408413673, |
| "grad_norm": 0.420953631401062, |
| "learning_rate": 9.227051338803656e-06, |
| "loss": 0.6692, |
| "step": 700 |
| }, |
| { |
| "epoch": 1.2611744084136722, |
| "grad_norm": 0.4463854432106018, |
| "learning_rate": 9.175802821066009e-06, |
| "loss": 0.6737, |
| "step": 720 |
| }, |
| { |
| "epoch": 1.2962313759859772, |
| "grad_norm": 0.44004735350608826, |
| "learning_rate": 9.12306141920832e-06, |
| "loss": 0.6673, |
| "step": 740 |
| }, |
| { |
| "epoch": 1.331288343558282, |
| "grad_norm": 0.42015475034713745, |
| "learning_rate": 9.068845988718906e-06, |
| "loss": 0.6676, |
| "step": 760 |
| }, |
| { |
| "epoch": 1.3663453111305872, |
| "grad_norm": 0.43683475255966187, |
| "learning_rate": 9.013175912063534e-06, |
| "loss": 0.6649, |
| "step": 780 |
| }, |
| { |
| "epoch": 1.4014022787028921, |
| "grad_norm": 0.4281805753707886, |
| "learning_rate": 8.956071091756036e-06, |
| "loss": 0.6658, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.4364592462751973, |
| "grad_norm": 0.4270734190940857, |
| "learning_rate": 8.89755194324299e-06, |
| "loss": 0.6646, |
| "step": 820 |
| }, |
| { |
| "epoch": 1.4715162138475022, |
| "grad_norm": 0.4163481593132019, |
| "learning_rate": 8.837639387605031e-06, |
| "loss": 0.6658, |
| "step": 840 |
| }, |
| { |
| "epoch": 1.5065731814198071, |
| "grad_norm": 0.45280900597572327, |
| "learning_rate": 8.776354844077389e-06, |
| "loss": 0.6592, |
| "step": 860 |
| }, |
| { |
| "epoch": 1.541630148992112, |
| "grad_norm": 0.40485361218452454, |
| "learning_rate": 8.713720222392338e-06, |
| "loss": 0.6579, |
| "step": 880 |
| }, |
| { |
| "epoch": 1.5766871165644172, |
| "grad_norm": 0.42039763927459717, |
| "learning_rate": 8.649757914946284e-06, |
| "loss": 0.6616, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.6117440841367223, |
| "grad_norm": 0.4760454595088959, |
| "learning_rate": 8.584490788794296e-06, |
| "loss": 0.6572, |
| "step": 920 |
| }, |
| { |
| "epoch": 1.6468010517090272, |
| "grad_norm": 0.43802690505981445, |
| "learning_rate": 8.517942177474943e-06, |
| "loss": 0.6548, |
| "step": 940 |
| }, |
| { |
| "epoch": 1.6818580192813322, |
| "grad_norm": 0.5002708435058594, |
| "learning_rate": 8.450135872668369e-06, |
| "loss": 0.6557, |
| "step": 960 |
| }, |
| { |
| "epoch": 1.716914986853637, |
| "grad_norm": 0.4160609543323517, |
| "learning_rate": 8.38109611569056e-06, |
| "loss": 0.6529, |
| "step": 980 |
| }, |
| { |
| "epoch": 1.751971954425942, |
| "grad_norm": 0.43179649114608765, |
| "learning_rate": 8.310847588826876e-06, |
| "loss": 0.6529, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.7870289219982471, |
| "grad_norm": 0.4322780668735504, |
| "learning_rate": 8.239415406507934e-06, |
| "loss": 0.6535, |
| "step": 1020 |
| }, |
| { |
| "epoch": 1.8220858895705523, |
| "grad_norm": 0.4642186462879181, |
| "learning_rate": 8.166825106330985e-06, |
| "loss": 0.649, |
| "step": 1040 |
| }, |
| { |
| "epoch": 1.8571428571428572, |
| "grad_norm": 0.42697349190711975, |
| "learning_rate": 8.093102639930013e-06, |
| "loss": 0.65, |
| "step": 1060 |
| }, |
| { |
| "epoch": 1.8921998247151621, |
| "grad_norm": 0.4486387372016907, |
| "learning_rate": 8.01827436369781e-06, |
| "loss": 0.6492, |
| "step": 1080 |
| }, |
| { |
| "epoch": 1.927256792287467, |
| "grad_norm": 0.42962825298309326, |
| "learning_rate": 7.942367029363351e-06, |
| "loss": 0.6518, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.962313759859772, |
| "grad_norm": 0.43645408749580383, |
| "learning_rate": 7.865407774427828e-06, |
| "loss": 0.6475, |
| "step": 1120 |
| }, |
| { |
| "epoch": 1.997370727432077, |
| "grad_norm": 0.4662039875984192, |
| "learning_rate": 7.787424112462758e-06, |
| "loss": 0.649, |
| "step": 1140 |
| }, |
| { |
| "epoch": 2.0315512708150747, |
| "grad_norm": 0.44152551889419556, |
| "learning_rate": 7.708443923273671e-06, |
| "loss": 0.6401, |
| "step": 1160 |
| }, |
| { |
| "epoch": 2.0666082383873796, |
| "grad_norm": 0.39342495799064636, |
| "learning_rate": 7.628495442932838e-06, |
| "loss": 0.6386, |
| "step": 1180 |
| }, |
| { |
| "epoch": 2.1016652059596845, |
| "grad_norm": 0.4356766641139984, |
| "learning_rate": 7.54760725368464e-06, |
| "loss": 0.6391, |
| "step": 1200 |
| }, |
| { |
| "epoch": 2.1367221735319895, |
| "grad_norm": 0.3897708058357239, |
| "learning_rate": 7.465808273727182e-06, |
| "loss": 0.6383, |
| "step": 1220 |
| }, |
| { |
| "epoch": 2.1717791411042944, |
| "grad_norm": 0.4168529510498047, |
| "learning_rate": 7.383127746873796e-06, |
| "loss": 0.6361, |
| "step": 1240 |
| }, |
| { |
| "epoch": 2.2068361086765993, |
| "grad_norm": 0.39462465047836304, |
| "learning_rate": 7.2995952320981356e-06, |
| "loss": 0.6371, |
| "step": 1260 |
| }, |
| { |
| "epoch": 2.2418930762489047, |
| "grad_norm": 0.42870041728019714, |
| "learning_rate": 7.215240592966603e-06, |
| "loss": 0.6318, |
| "step": 1280 |
| }, |
| { |
| "epoch": 2.2769500438212096, |
| "grad_norm": 0.46848800778388977, |
| "learning_rate": 7.130093986961868e-06, |
| "loss": 0.633, |
| "step": 1300 |
| }, |
| { |
| "epoch": 2.3120070113935145, |
| "grad_norm": 0.415912002325058, |
| "learning_rate": 7.044185854701321e-06, |
| "loss": 0.6367, |
| "step": 1320 |
| }, |
| { |
| "epoch": 2.3470639789658194, |
| "grad_norm": 0.4347931444644928, |
| "learning_rate": 6.957546909054304e-06, |
| "loss": 0.6374, |
| "step": 1340 |
| }, |
| { |
| "epoch": 2.3821209465381243, |
| "grad_norm": 0.4282444417476654, |
| "learning_rate": 6.870208124161998e-06, |
| "loss": 0.6353, |
| "step": 1360 |
| }, |
| { |
| "epoch": 2.4171779141104293, |
| "grad_norm": 0.43224233388900757, |
| "learning_rate": 6.78220072436392e-06, |
| "loss": 0.6348, |
| "step": 1380 |
| }, |
| { |
| "epoch": 2.4522348816827346, |
| "grad_norm": 0.4176190495491028, |
| "learning_rate": 6.693556173034953e-06, |
| "loss": 0.633, |
| "step": 1400 |
| }, |
| { |
| "epoch": 2.4872918492550395, |
| "grad_norm": 0.4181615710258484, |
| "learning_rate": 6.6043061613369356e-06, |
| "loss": 0.6324, |
| "step": 1420 |
| }, |
| { |
| "epoch": 2.5223488168273445, |
| "grad_norm": 0.38148173689842224, |
| "learning_rate": 6.514482596888807e-06, |
| "loss": 0.6301, |
| "step": 1440 |
| }, |
| { |
| "epoch": 2.5574057843996494, |
| "grad_norm": 0.41031816601753235, |
| "learning_rate": 6.424117592359367e-06, |
| "loss": 0.6332, |
| "step": 1460 |
| }, |
| { |
| "epoch": 2.5924627519719543, |
| "grad_norm": 0.4005562961101532, |
| "learning_rate": 6.333243453986734e-06, |
| "loss": 0.6265, |
| "step": 1480 |
| }, |
| { |
| "epoch": 2.6275197195442592, |
| "grad_norm": 0.3940238058567047, |
| "learning_rate": 6.241892670028595e-06, |
| "loss": 0.6315, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.662576687116564, |
| "grad_norm": 0.4001730680465698, |
| "learning_rate": 6.150097899147384e-06, |
| "loss": 0.6299, |
| "step": 1520 |
| }, |
| { |
| "epoch": 2.6976336546888695, |
| "grad_norm": 0.3857872188091278, |
| "learning_rate": 6.057891958734538e-06, |
| "loss": 0.6304, |
| "step": 1540 |
| }, |
| { |
| "epoch": 2.7326906222611744, |
| "grad_norm": 0.4169263243675232, |
| "learning_rate": 5.965307813178015e-06, |
| "loss": 0.6315, |
| "step": 1560 |
| }, |
| { |
| "epoch": 2.7677475898334793, |
| "grad_norm": 0.4010975658893585, |
| "learning_rate": 5.872378562077241e-06, |
| "loss": 0.6297, |
| "step": 1580 |
| }, |
| { |
| "epoch": 2.8028045574057843, |
| "grad_norm": 0.4302142262458801, |
| "learning_rate": 5.779137428409738e-06, |
| "loss": 0.6302, |
| "step": 1600 |
| }, |
| { |
| "epoch": 2.8378615249780896, |
| "grad_norm": 0.39000585675239563, |
| "learning_rate": 5.685617746653629e-06, |
| "loss": 0.6312, |
| "step": 1620 |
| }, |
| { |
| "epoch": 2.8729184925503946, |
| "grad_norm": 0.4292212724685669, |
| "learning_rate": 5.591852950870287e-06, |
| "loss": 0.6312, |
| "step": 1640 |
| }, |
| { |
| "epoch": 2.9079754601226995, |
| "grad_norm": 0.3838886320590973, |
| "learning_rate": 5.497876562751384e-06, |
| "loss": 0.6302, |
| "step": 1660 |
| }, |
| { |
| "epoch": 2.9430324276950044, |
| "grad_norm": 0.36835259199142456, |
| "learning_rate": 5.403722179634602e-06, |
| "loss": 0.6292, |
| "step": 1680 |
| }, |
| { |
| "epoch": 2.9780893952673093, |
| "grad_norm": 0.3884848654270172, |
| "learning_rate": 5.309423462492314e-06, |
| "loss": 0.6261, |
| "step": 1700 |
| }, |
| { |
| "epoch": 3.0122699386503067, |
| "grad_norm": 0.3762246072292328, |
| "learning_rate": 5.215014123897504e-06, |
| "loss": 0.6202, |
| "step": 1720 |
| }, |
| { |
| "epoch": 3.0473269062226116, |
| "grad_norm": 0.38138872385025024, |
| "learning_rate": 5.120527915971235e-06, |
| "loss": 0.6205, |
| "step": 1740 |
| }, |
| { |
| "epoch": 3.0823838737949165, |
| "grad_norm": 0.38698920607566833, |
| "learning_rate": 5.0259986183160006e-06, |
| "loss": 0.6186, |
| "step": 1760 |
| }, |
| { |
| "epoch": 3.117440841367222, |
| "grad_norm": 0.378830224275589, |
| "learning_rate": 4.931460025939226e-06, |
| "loss": 0.6214, |
| "step": 1780 |
| }, |
| { |
| "epoch": 3.152497808939527, |
| "grad_norm": 0.3751004934310913, |
| "learning_rate": 4.836945937171279e-06, |
| "loss": 0.6209, |
| "step": 1800 |
| }, |
| { |
| "epoch": 3.1875547765118317, |
| "grad_norm": 0.3829745054244995, |
| "learning_rate": 4.742490141582279e-06, |
| "loss": 0.6213, |
| "step": 1820 |
| }, |
| { |
| "epoch": 3.2226117440841366, |
| "grad_norm": 0.4228389859199524, |
| "learning_rate": 4.648126407902058e-06, |
| "loss": 0.6193, |
| "step": 1840 |
| }, |
| { |
| "epoch": 3.2576687116564416, |
| "grad_norm": 0.367960125207901, |
| "learning_rate": 4.553888471947546e-06, |
| "loss": 0.6198, |
| "step": 1860 |
| }, |
| { |
| "epoch": 3.292725679228747, |
| "grad_norm": 0.39815646409988403, |
| "learning_rate": 4.4598100245619505e-06, |
| "loss": 0.6187, |
| "step": 1880 |
| }, |
| { |
| "epoch": 3.327782646801052, |
| "grad_norm": 0.3625248074531555, |
| "learning_rate": 4.3659246995699845e-06, |
| "loss": 0.6176, |
| "step": 1900 |
| }, |
| { |
| "epoch": 3.3628396143733568, |
| "grad_norm": 0.37671083211898804, |
| "learning_rate": 4.2722660617535105e-06, |
| "loss": 0.6182, |
| "step": 1920 |
| }, |
| { |
| "epoch": 3.3978965819456617, |
| "grad_norm": 0.3727245032787323, |
| "learning_rate": 4.178867594851849e-06, |
| "loss": 0.616, |
| "step": 1940 |
| }, |
| { |
| "epoch": 3.4329535495179666, |
| "grad_norm": 0.361914724111557, |
| "learning_rate": 4.085762689591054e-06, |
| "loss": 0.6157, |
| "step": 1960 |
| }, |
| { |
| "epoch": 3.4680105170902715, |
| "grad_norm": 0.3587988018989563, |
| "learning_rate": 3.992984631746469e-06, |
| "loss": 0.6188, |
| "step": 1980 |
| }, |
| { |
| "epoch": 3.5030674846625764, |
| "grad_norm": 0.3729381561279297, |
| "learning_rate": 3.9005665902427695e-06, |
| "loss": 0.6208, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.538124452234882, |
| "grad_norm": 0.35941800475120544, |
| "learning_rate": 3.8085416052958107e-06, |
| "loss": 0.6183, |
| "step": 2020 |
| }, |
| { |
| "epoch": 3.5731814198071867, |
| "grad_norm": 0.35061511397361755, |
| "learning_rate": 3.7169425766004653e-06, |
| "loss": 0.6148, |
| "step": 2040 |
| }, |
| { |
| "epoch": 3.6082383873794917, |
| "grad_norm": 0.36660343408584595, |
| "learning_rate": 3.6258022515687215e-06, |
| "loss": 0.6186, |
| "step": 2060 |
| }, |
| { |
| "epoch": 3.6432953549517966, |
| "grad_norm": 0.4035237729549408, |
| "learning_rate": 3.5351532136222012e-06, |
| "loss": 0.616, |
| "step": 2080 |
| }, |
| { |
| "epoch": 3.678352322524102, |
| "grad_norm": 0.3646794855594635, |
| "learning_rate": 3.445027870543323e-06, |
| "loss": 0.6145, |
| "step": 2100 |
| }, |
| { |
| "epoch": 3.713409290096407, |
| "grad_norm": 0.38012927770614624, |
| "learning_rate": 3.3554584428892488e-06, |
| "loss": 0.6181, |
| "step": 2120 |
| }, |
| { |
| "epoch": 3.7484662576687118, |
| "grad_norm": 0.36509743332862854, |
| "learning_rate": 3.2664769524727712e-06, |
| "loss": 0.616, |
| "step": 2140 |
| }, |
| { |
| "epoch": 3.7835232252410167, |
| "grad_norm": 0.38520926237106323, |
| "learning_rate": 3.178115210914242e-06, |
| "loss": 0.6169, |
| "step": 2160 |
| }, |
| { |
| "epoch": 3.8185801928133216, |
| "grad_norm": 0.3496517837047577, |
| "learning_rate": 3.0904048082686655e-06, |
| "loss": 0.6158, |
| "step": 2180 |
| }, |
| { |
| "epoch": 3.8536371603856265, |
| "grad_norm": 0.37065935134887695, |
| "learning_rate": 3.00337710173198e-06, |
| "loss": 0.6162, |
| "step": 2200 |
| }, |
| { |
| "epoch": 3.8886941279579315, |
| "grad_norm": 0.339855819940567, |
| "learning_rate": 2.9170632044306137e-06, |
| "loss": 0.6156, |
| "step": 2220 |
| }, |
| { |
| "epoch": 3.9237510955302364, |
| "grad_norm": 0.3591175675392151, |
| "learning_rate": 2.8314939742982673e-06, |
| "loss": 0.6144, |
| "step": 2240 |
| }, |
| { |
| "epoch": 3.9588080631025417, |
| "grad_norm": 0.3600142300128937, |
| "learning_rate": 2.746700003043964e-06, |
| "loss": 0.6195, |
| "step": 2260 |
| }, |
| { |
| "epoch": 3.9938650306748467, |
| "grad_norm": 0.3738831579685211, |
| "learning_rate": 2.662711605215248e-06, |
| "loss": 0.6135, |
| "step": 2280 |
| }, |
| { |
| "epoch": 4.028045574057844, |
| "grad_norm": 0.3281383514404297, |
| "learning_rate": 2.579558807360489e-06, |
| "loss": 0.6104, |
| "step": 2300 |
| }, |
| { |
| "epoch": 4.063102541630149, |
| "grad_norm": 0.3486866354942322, |
| "learning_rate": 2.4972713372941406e-06, |
| "loss": 0.6095, |
| "step": 2320 |
| }, |
| { |
| "epoch": 4.098159509202454, |
| "grad_norm": 0.3392680287361145, |
| "learning_rate": 2.4158786134687966e-06, |
| "loss": 0.6108, |
| "step": 2340 |
| }, |
| { |
| "epoch": 4.133216476774759, |
| "grad_norm": 0.3901905417442322, |
| "learning_rate": 2.3354097344578565e-06, |
| "loss": 0.6123, |
| "step": 2360 |
| }, |
| { |
| "epoch": 4.168273444347064, |
| "grad_norm": 0.33477798104286194, |
| "learning_rate": 2.25589346855254e-06, |
| "loss": 0.6079, |
| "step": 2380 |
| }, |
| { |
| "epoch": 4.203330411919369, |
| "grad_norm": 0.35764065384864807, |
| "learning_rate": 2.1773582434769854e-06, |
| "loss": 0.6066, |
| "step": 2400 |
| }, |
| { |
| "epoch": 4.238387379491674, |
| "grad_norm": 0.34379830956459045, |
| "learning_rate": 2.0998321362251036e-06, |
| "loss": 0.608, |
| "step": 2420 |
| }, |
| { |
| "epoch": 4.273444347063979, |
| "grad_norm": 0.3394622206687927, |
| "learning_rate": 2.023342863022819e-06, |
| "loss": 0.6142, |
| "step": 2440 |
| }, |
| { |
| "epoch": 4.308501314636284, |
| "grad_norm": 0.35136643052101135, |
| "learning_rate": 1.94791776941929e-06, |
| "loss": 0.6115, |
| "step": 2460 |
| }, |
| { |
| "epoch": 4.343558282208589, |
| "grad_norm": 0.3298383355140686, |
| "learning_rate": 1.873583820510647e-06, |
| "loss": 0.6089, |
| "step": 2480 |
| }, |
| { |
| "epoch": 4.378615249780894, |
| "grad_norm": 0.3390386402606964, |
| "learning_rate": 1.8003675912997487e-06, |
| "loss": 0.612, |
| "step": 2500 |
| }, |
| { |
| "epoch": 4.413672217353199, |
| "grad_norm": 0.3399540185928345, |
| "learning_rate": 1.7282952571953987e-06, |
| "loss": 0.6123, |
| "step": 2520 |
| }, |
| { |
| "epoch": 4.448729184925504, |
| "grad_norm": 0.3504091203212738, |
| "learning_rate": 1.657392584654412e-06, |
| "loss": 0.6078, |
| "step": 2540 |
| }, |
| { |
| "epoch": 4.483786152497809, |
| "grad_norm": 0.3221462070941925, |
| "learning_rate": 1.587684921969912e-06, |
| "loss": 0.6142, |
| "step": 2560 |
| }, |
| { |
| "epoch": 4.518843120070114, |
| "grad_norm": 0.34956008195877075, |
| "learning_rate": 1.5191971902090797e-06, |
| "loss": 0.6085, |
| "step": 2580 |
| }, |
| { |
| "epoch": 4.553900087642419, |
| "grad_norm": 0.34091153740882874, |
| "learning_rate": 1.4519538743036927e-06, |
| "loss": 0.6117, |
| "step": 2600 |
| }, |
| { |
| "epoch": 4.588957055214724, |
| "grad_norm": 0.34645262360572815, |
| "learning_rate": 1.385979014296533e-06, |
| "loss": 0.6076, |
| "step": 2620 |
| }, |
| { |
| "epoch": 4.624014022787029, |
| "grad_norm": 0.3348851799964905, |
| "learning_rate": 1.3212961967468985e-06, |
| "loss": 0.6116, |
| "step": 2640 |
| }, |
| { |
| "epoch": 4.659070990359334, |
| "grad_norm": 0.32889384031295776, |
| "learning_rate": 1.2579285462981855e-06, |
| "loss": 0.6129, |
| "step": 2660 |
| }, |
| { |
| "epoch": 4.694127957931639, |
| "grad_norm": 0.3461220860481262, |
| "learning_rate": 1.195898717410664e-06, |
| "loss": 0.6106, |
| "step": 2680 |
| }, |
| { |
| "epoch": 4.729184925503944, |
| "grad_norm": 0.319545716047287, |
| "learning_rate": 1.1352288862622968e-06, |
| "loss": 0.6068, |
| "step": 2700 |
| }, |
| { |
| "epoch": 4.764241893076249, |
| "grad_norm": 0.3272876739501953, |
| "learning_rate": 1.075940742820588e-06, |
| "loss": 0.6082, |
| "step": 2720 |
| }, |
| { |
| "epoch": 4.799298860648554, |
| "grad_norm": 0.3398887515068054, |
| "learning_rate": 1.0180554830882333e-06, |
| "loss": 0.6069, |
| "step": 2740 |
| }, |
| { |
| "epoch": 4.8343558282208585, |
| "grad_norm": 0.32064250111579895, |
| "learning_rate": 9.615938015253723e-07, |
| "loss": 0.6086, |
| "step": 2760 |
| }, |
| { |
| "epoch": 4.869412795793163, |
| "grad_norm": 0.32411012053489685, |
| "learning_rate": 9.065758836511556e-07, |
| "loss": 0.6083, |
| "step": 2780 |
| }, |
| { |
| "epoch": 4.904469763365469, |
| "grad_norm": 0.31640660762786865, |
| "learning_rate": 8.53021398827239e-07, |
| "loss": 0.6094, |
| "step": 2800 |
| }, |
| { |
| "epoch": 4.939526730937774, |
| "grad_norm": 0.3243345320224762, |
| "learning_rate": 8.009494932258427e-07, |
| "loss": 0.6104, |
| "step": 2820 |
| }, |
| { |
| "epoch": 4.974583698510079, |
| "grad_norm": 0.31760430335998535, |
| "learning_rate": 7.503787829848191e-07, |
| "loss": 0.6078, |
| "step": 2840 |
| }, |
| { |
| "epoch": 5.0087642418930765, |
| "grad_norm": 0.32336053252220154, |
| "learning_rate": 7.013273475522392e-07, |
| "loss": 0.609, |
| "step": 2860 |
| }, |
| { |
| "epoch": 5.043821209465381, |
| "grad_norm": 0.31958338618278503, |
| "learning_rate": 6.53812723222838e-07, |
| "loss": 0.6073, |
| "step": 2880 |
| }, |
| { |
| "epoch": 5.078878177037686, |
| "grad_norm": 0.319396436214447, |
| "learning_rate": 6.078518968686426e-07, |
| "loss": 0.6096, |
| "step": 2900 |
| }, |
| { |
| "epoch": 5.113935144609991, |
| "grad_norm": 0.33287620544433594, |
| "learning_rate": 5.634612998660249e-07, |
| "loss": 0.6011, |
| "step": 2920 |
| }, |
| { |
| "epoch": 5.148992112182296, |
| "grad_norm": 0.32154789566993713, |
| "learning_rate": 5.206568022213482e-07, |
| "loss": 0.6081, |
| "step": 2940 |
| }, |
| { |
| "epoch": 5.184049079754601, |
| "grad_norm": 0.3217925727367401, |
| "learning_rate": 4.794537068973093e-07, |
| "loss": 0.607, |
| "step": 2960 |
| }, |
| { |
| "epoch": 5.219106047326906, |
| "grad_norm": 0.3143027722835541, |
| "learning_rate": 4.398667443420029e-07, |
| "loss": 0.6065, |
| "step": 2980 |
| }, |
| { |
| "epoch": 5.254163014899211, |
| "grad_norm": 0.3175284266471863, |
| "learning_rate": 4.019100672226617e-07, |
| "loss": 0.6056, |
| "step": 3000 |
| }, |
| { |
| "epoch": 5.289219982471516, |
| "grad_norm": 0.32859277725219727, |
| "learning_rate": 3.65597245365964e-07, |
| "loss": 0.6063, |
| "step": 3020 |
| }, |
| { |
| "epoch": 5.324276950043822, |
| "grad_norm": 0.3200486898422241, |
| "learning_rate": 3.3094126090670477e-07, |
| "loss": 0.6074, |
| "step": 3040 |
| }, |
| { |
| "epoch": 5.3593339176161265, |
| "grad_norm": 0.3193705976009369, |
| "learning_rate": 2.9795450364657865e-07, |
| "loss": 0.607, |
| "step": 3060 |
| }, |
| { |
| "epoch": 5.3943908851884315, |
| "grad_norm": 0.32681772112846375, |
| "learning_rate": 2.6664876662471697e-07, |
| "loss": 0.6053, |
| "step": 3080 |
| }, |
| { |
| "epoch": 5.429447852760736, |
| "grad_norm": 0.31692323088645935, |
| "learning_rate": 2.370352419015892e-07, |
| "loss": 0.608, |
| "step": 3100 |
| }, |
| { |
| "epoch": 5.464504820333041, |
| "grad_norm": 0.32039353251457214, |
| "learning_rate": 2.091245165577349e-07, |
| "loss": 0.6079, |
| "step": 3120 |
| }, |
| { |
| "epoch": 5.499561787905346, |
| "grad_norm": 0.31619492173194885, |
| "learning_rate": 1.8292656890880722e-07, |
| "loss": 0.6045, |
| "step": 3140 |
| }, |
| { |
| "epoch": 5.534618755477651, |
| "grad_norm": 0.31279516220092773, |
| "learning_rate": 1.5845076493823331e-07, |
| "loss": 0.6066, |
| "step": 3160 |
| }, |
| { |
| "epoch": 5.569675723049956, |
| "grad_norm": 0.3182794451713562, |
| "learning_rate": 1.3570585494880328e-07, |
| "loss": 0.6085, |
| "step": 3180 |
| }, |
| { |
| "epoch": 5.604732690622261, |
| "grad_norm": 0.31176891922950745, |
| "learning_rate": 1.1469997043436154e-07, |
| "loss": 0.6048, |
| "step": 3200 |
| }, |
| { |
| "epoch": 5.639789658194566, |
| "grad_norm": 0.31051337718963623, |
| "learning_rate": 9.544062117273045e-08, |
| "loss": 0.607, |
| "step": 3220 |
| }, |
| { |
| "epoch": 5.674846625766871, |
| "grad_norm": 0.31119874119758606, |
| "learning_rate": 7.793469254090524e-08, |
| "loss": 0.6054, |
| "step": 3240 |
| }, |
| { |
| "epoch": 5.709903593339176, |
| "grad_norm": 0.3019949793815613, |
| "learning_rate": 6.218844305346916e-08, |
| "loss": 0.6074, |
| "step": 3260 |
| }, |
| { |
| "epoch": 5.7449605609114816, |
| "grad_norm": 0.30909818410873413, |
| "learning_rate": 4.820750212513048e-08, |
| "loss": 0.6085, |
| "step": 3280 |
| }, |
| { |
| "epoch": 5.7800175284837865, |
| "grad_norm": 0.30439695715904236, |
| "learning_rate": 3.599686805815128e-08, |
| "loss": 0.6101, |
| "step": 3300 |
| }, |
| { |
| "epoch": 5.815074496056091, |
| "grad_norm": 0.3216073215007782, |
| "learning_rate": 2.5560906255420737e-08, |
| "loss": 0.6092, |
| "step": 3320 |
| }, |
| { |
| "epoch": 5.850131463628396, |
| "grad_norm": 0.3213990330696106, |
| "learning_rate": 1.6903347659781856e-08, |
| "loss": 0.6071, |
| "step": 3340 |
| }, |
| { |
| "epoch": 5.885188431200701, |
| "grad_norm": 0.3153989613056183, |
| "learning_rate": 1.0027287420192322e-08, |
| "loss": 0.6074, |
| "step": 3360 |
| }, |
| { |
| "epoch": 5.920245398773006, |
| "grad_norm": 0.308703750371933, |
| "learning_rate": 4.935183785180209e-09, |
| "loss": 0.6058, |
| "step": 3380 |
| }, |
| { |
| "epoch": 5.955302366345311, |
| "grad_norm": 0.3108364939689636, |
| "learning_rate": 1.6288572240014123e-09, |
| "loss": 0.6038, |
| "step": 3400 |
| }, |
| { |
| "epoch": 5.990359333917616, |
| "grad_norm": 0.313579261302948, |
| "learning_rate": 1.0948977580638176e-10, |
| "loss": 0.6079, |
| "step": 3420 |
| }, |
| { |
| "epoch": 6.0, |
| "step": 3426, |
| "total_flos": 9.485754681343869e+18, |
| "train_loss": 0.6827397172519218, |
| "train_runtime": 47545.1809, |
| "train_samples_per_second": 4.607, |
| "train_steps_per_second": 0.072 |
| } |
| ], |
| "logging_steps": 20, |
| "max_steps": 3426, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 6, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 9.485754681343869e+18, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
