| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.040716214499532576, | |
| "eval_steps": 500, | |
| "global_step": 12500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0003257297159962606, | |
| "grad_norm": 2.2308592796325684, | |
| "learning_rate": 4.99853416853153e-05, | |
| "loss": 1.4483, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0006514594319925212, | |
| "grad_norm": 2.3997225761413574, | |
| "learning_rate": 4.996905466899897e-05, | |
| "loss": 1.3276, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0009771891479887819, | |
| "grad_norm": 1.4687339067459106, | |
| "learning_rate": 4.995276765268264e-05, | |
| "loss": 1.3394, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0013029188639850425, | |
| "grad_norm": 0.6583470702171326, | |
| "learning_rate": 4.993648063636631e-05, | |
| "loss": 1.3245, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0016286485799813031, | |
| "grad_norm": 1.6252340078353882, | |
| "learning_rate": 4.992019362004997e-05, | |
| "loss": 1.3249, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0019543782959775637, | |
| "grad_norm": 2.0806777477264404, | |
| "learning_rate": 4.9903906603733634e-05, | |
| "loss": 1.32, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.002280108011973824, | |
| "grad_norm": 1.376539707183838, | |
| "learning_rate": 4.988761958741731e-05, | |
| "loss": 1.3133, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.002605837727970085, | |
| "grad_norm": 2.234644889831543, | |
| "learning_rate": 4.987133257110097e-05, | |
| "loss": 1.3179, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.0029315674439663454, | |
| "grad_norm": 1.4599684476852417, | |
| "learning_rate": 4.985504555478464e-05, | |
| "loss": 1.3097, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.0032572971599626062, | |
| "grad_norm": 1.7078094482421875, | |
| "learning_rate": 4.9838758538468304e-05, | |
| "loss": 1.3083, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.0035830268759588666, | |
| "grad_norm": 0.6953567266464233, | |
| "learning_rate": 4.9822471522151976e-05, | |
| "loss": 1.3075, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.0039087565919551275, | |
| "grad_norm": 1.225602626800537, | |
| "learning_rate": 4.980618450583564e-05, | |
| "loss": 1.3054, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.004234486307951388, | |
| "grad_norm": 1.3010519742965698, | |
| "learning_rate": 4.978989748951931e-05, | |
| "loss": 1.3066, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.004560216023947648, | |
| "grad_norm": 0.6475724577903748, | |
| "learning_rate": 4.9773610473202974e-05, | |
| "loss": 1.3109, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.004885945739943909, | |
| "grad_norm": 1.046614646911621, | |
| "learning_rate": 4.975732345688664e-05, | |
| "loss": 1.3074, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.00521167545594017, | |
| "grad_norm": 1.113573670387268, | |
| "learning_rate": 4.974103644057031e-05, | |
| "loss": 1.3083, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.005537405171936431, | |
| "grad_norm": 1.4273550510406494, | |
| "learning_rate": 4.972474942425398e-05, | |
| "loss": 1.3018, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.005863134887932691, | |
| "grad_norm": 0.5519908666610718, | |
| "learning_rate": 4.970846240793764e-05, | |
| "loss": 1.2945, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.006188864603928952, | |
| "grad_norm": 0.6653416156768799, | |
| "learning_rate": 4.969217539162131e-05, | |
| "loss": 1.3004, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.0065145943199252125, | |
| "grad_norm": 0.732170581817627, | |
| "learning_rate": 4.9675888375304975e-05, | |
| "loss": 1.3014, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.006840324035921473, | |
| "grad_norm": 0.405608594417572, | |
| "learning_rate": 4.965960135898865e-05, | |
| "loss": 1.2939, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.007166053751917733, | |
| "grad_norm": 0.9849847555160522, | |
| "learning_rate": 4.9643314342672306e-05, | |
| "loss": 1.2922, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.007491783467913994, | |
| "grad_norm": 0.7152832746505737, | |
| "learning_rate": 4.962702732635598e-05, | |
| "loss": 1.2905, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.007817513183910255, | |
| "grad_norm": 1.1164734363555908, | |
| "learning_rate": 4.9610740310039644e-05, | |
| "loss": 1.3024, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.008143242899906516, | |
| "grad_norm": 0.574243426322937, | |
| "learning_rate": 4.959445329372332e-05, | |
| "loss": 1.2944, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.008468972615902777, | |
| "grad_norm": 0.6976324319839478, | |
| "learning_rate": 4.9578166277406976e-05, | |
| "loss": 1.2939, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.008794702331899037, | |
| "grad_norm": 0.4648737609386444, | |
| "learning_rate": 4.956187926109064e-05, | |
| "loss": 1.2841, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.009120432047895297, | |
| "grad_norm": 1.189271092414856, | |
| "learning_rate": 4.9545592244774314e-05, | |
| "loss": 1.294, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.009446161763891557, | |
| "grad_norm": 0.6437670588493347, | |
| "learning_rate": 4.952930522845798e-05, | |
| "loss": 1.2882, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.009771891479887818, | |
| "grad_norm": 1.591304898262024, | |
| "learning_rate": 4.9513018212141646e-05, | |
| "loss": 1.2805, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.010097621195884079, | |
| "grad_norm": 0.2836475670337677, | |
| "learning_rate": 4.949673119582531e-05, | |
| "loss": 1.2802, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.01042335091188034, | |
| "grad_norm": 1.304417610168457, | |
| "learning_rate": 4.9480444179508984e-05, | |
| "loss": 1.2833, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.0107490806278766, | |
| "grad_norm": 0.27579864859580994, | |
| "learning_rate": 4.946415716319265e-05, | |
| "loss": 1.2852, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.011074810343872862, | |
| "grad_norm": 1.1080585718154907, | |
| "learning_rate": 4.9447870146876315e-05, | |
| "loss": 1.289, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.011400540059869122, | |
| "grad_norm": 0.2783690392971039, | |
| "learning_rate": 4.943158313055998e-05, | |
| "loss": 1.2885, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.011726269775865382, | |
| "grad_norm": 0.6603112816810608, | |
| "learning_rate": 4.941529611424365e-05, | |
| "loss": 1.2882, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.012051999491861642, | |
| "grad_norm": 0.9498095512390137, | |
| "learning_rate": 4.939900909792732e-05, | |
| "loss": 1.2835, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.012377729207857903, | |
| "grad_norm": 0.5274548530578613, | |
| "learning_rate": 4.9382722081610985e-05, | |
| "loss": 1.279, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.012703458923854164, | |
| "grad_norm": 0.5299821496009827, | |
| "learning_rate": 4.936643506529465e-05, | |
| "loss": 1.2879, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.013029188639850425, | |
| "grad_norm": 1.0898863077163696, | |
| "learning_rate": 4.9350148048978316e-05, | |
| "loss": 1.2913, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.013354918355846686, | |
| "grad_norm": 0.6892501711845398, | |
| "learning_rate": 4.933386103266198e-05, | |
| "loss": 1.2835, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.013680648071842947, | |
| "grad_norm": 0.9103847146034241, | |
| "learning_rate": 4.9317574016345655e-05, | |
| "loss": 1.2876, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.014006377787839207, | |
| "grad_norm": 0.8750960826873779, | |
| "learning_rate": 4.9301287000029314e-05, | |
| "loss": 1.2761, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.014332107503835467, | |
| "grad_norm": 1.7296843528747559, | |
| "learning_rate": 4.9284999983712986e-05, | |
| "loss": 1.2825, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.014657837219831727, | |
| "grad_norm": 0.7019387483596802, | |
| "learning_rate": 4.926871296739665e-05, | |
| "loss": 1.2774, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.014983566935827988, | |
| "grad_norm": 0.9353660345077515, | |
| "learning_rate": 4.9252425951080324e-05, | |
| "loss": 1.2701, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.015309296651824249, | |
| "grad_norm": 0.7081932425498962, | |
| "learning_rate": 4.923613893476399e-05, | |
| "loss": 1.276, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.01563502636782051, | |
| "grad_norm": 0.8366962671279907, | |
| "learning_rate": 4.9219851918447656e-05, | |
| "loss": 1.2767, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.01596075608381677, | |
| "grad_norm": 1.765871286392212, | |
| "learning_rate": 4.920356490213132e-05, | |
| "loss": 1.2617, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.01628648579981303, | |
| "grad_norm": 0.2926379442214966, | |
| "learning_rate": 4.918727788581499e-05, | |
| "loss": 1.2762, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.01661221551580929, | |
| "grad_norm": 1.1176525354385376, | |
| "learning_rate": 4.917099086949866e-05, | |
| "loss": 1.2647, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.016937945231805553, | |
| "grad_norm": 0.384264200925827, | |
| "learning_rate": 4.915470385318232e-05, | |
| "loss": 1.2628, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.017263674947801812, | |
| "grad_norm": 1.5339140892028809, | |
| "learning_rate": 4.913841683686599e-05, | |
| "loss": 1.2692, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.017589404663798075, | |
| "grad_norm": 1.2026703357696533, | |
| "learning_rate": 4.912212982054966e-05, | |
| "loss": 1.2618, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.017915134379794334, | |
| "grad_norm": 0.6754997968673706, | |
| "learning_rate": 4.910584280423333e-05, | |
| "loss": 1.2495, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.018240864095790593, | |
| "grad_norm": 0.8240428566932678, | |
| "learning_rate": 4.908955578791699e-05, | |
| "loss": 1.2498, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.018566593811786856, | |
| "grad_norm": 0.6363087892532349, | |
| "learning_rate": 4.9073268771600654e-05, | |
| "loss": 1.2514, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.018892323527783115, | |
| "grad_norm": 1.393833875656128, | |
| "learning_rate": 4.905698175528433e-05, | |
| "loss": 1.2509, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.019218053243779377, | |
| "grad_norm": 0.6422170996665955, | |
| "learning_rate": 4.904069473896799e-05, | |
| "loss": 1.2405, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.019543782959775637, | |
| "grad_norm": 0.7575420141220093, | |
| "learning_rate": 4.902440772265166e-05, | |
| "loss": 1.2241, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.0198695126757719, | |
| "grad_norm": 0.7148196697235107, | |
| "learning_rate": 4.9008120706335324e-05, | |
| "loss": 1.2372, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.020195242391768158, | |
| "grad_norm": 1.1207329034805298, | |
| "learning_rate": 4.8991833690018996e-05, | |
| "loss": 1.2372, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.02052097210776442, | |
| "grad_norm": 1.3915568590164185, | |
| "learning_rate": 4.897554667370266e-05, | |
| "loss": 1.2129, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.02084670182376068, | |
| "grad_norm": 0.8674553036689758, | |
| "learning_rate": 4.895925965738633e-05, | |
| "loss": 1.2262, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.02117243153975694, | |
| "grad_norm": 0.7640644311904907, | |
| "learning_rate": 4.8942972641069994e-05, | |
| "loss": 1.1998, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.0214981612557532, | |
| "grad_norm": 0.7928606271743774, | |
| "learning_rate": 4.892668562475366e-05, | |
| "loss": 1.1776, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.02182389097174946, | |
| "grad_norm": 1.1644946336746216, | |
| "learning_rate": 4.891039860843733e-05, | |
| "loss": 1.1916, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.022149620687745723, | |
| "grad_norm": 1.1310213804244995, | |
| "learning_rate": 4.8894111592121e-05, | |
| "loss": 1.1786, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.022475350403741982, | |
| "grad_norm": 1.3858141899108887, | |
| "learning_rate": 4.887782457580466e-05, | |
| "loss": 1.1728, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.022801080119738245, | |
| "grad_norm": 3.814767360687256, | |
| "learning_rate": 4.886153755948833e-05, | |
| "loss": 1.1384, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.023126809835734504, | |
| "grad_norm": 1.2411885261535645, | |
| "learning_rate": 4.8845250543171995e-05, | |
| "loss": 1.1588, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.023452539551730763, | |
| "grad_norm": 1.4492881298065186, | |
| "learning_rate": 4.882896352685567e-05, | |
| "loss": 1.1266, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.023778269267727026, | |
| "grad_norm": 0.8389878869056702, | |
| "learning_rate": 4.8812676510539326e-05, | |
| "loss": 1.1446, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.024103998983723285, | |
| "grad_norm": 0.33955487608909607, | |
| "learning_rate": 4.8796389494223e-05, | |
| "loss": 1.1111, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.024429728699719547, | |
| "grad_norm": 0.7004753351211548, | |
| "learning_rate": 4.8780102477906664e-05, | |
| "loss": 1.0954, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.024755458415715807, | |
| "grad_norm": 0.7213209271430969, | |
| "learning_rate": 4.876381546159034e-05, | |
| "loss": 1.1123, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.02508118813171207, | |
| "grad_norm": 0.960991382598877, | |
| "learning_rate": 4.8747528445273996e-05, | |
| "loss": 1.0982, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.025406917847708328, | |
| "grad_norm": 0.6955804228782654, | |
| "learning_rate": 4.873124142895766e-05, | |
| "loss": 1.0827, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.02573264756370459, | |
| "grad_norm": 0.47498619556427, | |
| "learning_rate": 4.8714954412641334e-05, | |
| "loss": 1.1043, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.02605837727970085, | |
| "grad_norm": 0.304063618183136, | |
| "learning_rate": 4.8698667396325e-05, | |
| "loss": 1.0699, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.02638410699569711, | |
| "grad_norm": 0.9996088743209839, | |
| "learning_rate": 4.8682380380008666e-05, | |
| "loss": 1.0697, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.02670983671169337, | |
| "grad_norm": 0.5986392498016357, | |
| "learning_rate": 4.866609336369233e-05, | |
| "loss": 1.0733, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.02703556642768963, | |
| "grad_norm": 0.41347017884254456, | |
| "learning_rate": 4.8649806347376004e-05, | |
| "loss": 1.0643, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.027361296143685893, | |
| "grad_norm": 0.3976612687110901, | |
| "learning_rate": 4.863351933105967e-05, | |
| "loss": 1.0401, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.027687025859682152, | |
| "grad_norm": 1.1716387271881104, | |
| "learning_rate": 4.8617232314743335e-05, | |
| "loss": 1.0298, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.028012755575678415, | |
| "grad_norm": 0.7384105324745178, | |
| "learning_rate": 4.8600945298427e-05, | |
| "loss": 1.0223, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.028338485291674674, | |
| "grad_norm": 0.517280638217926, | |
| "learning_rate": 4.858465828211067e-05, | |
| "loss": 1.0445, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.028664215007670933, | |
| "grad_norm": 0.7129126787185669, | |
| "learning_rate": 4.856837126579434e-05, | |
| "loss": 1.0508, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.028989944723667196, | |
| "grad_norm": 0.35596320033073425, | |
| "learning_rate": 4.8552084249478005e-05, | |
| "loss": 1.0296, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.029315674439663455, | |
| "grad_norm": 0.9362590909004211, | |
| "learning_rate": 4.853579723316167e-05, | |
| "loss": 1.0785, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.029641404155659717, | |
| "grad_norm": 0.8223775625228882, | |
| "learning_rate": 4.8519510216845336e-05, | |
| "loss": 1.043, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.029967133871655977, | |
| "grad_norm": 0.7149192690849304, | |
| "learning_rate": 4.8503223200529e-05, | |
| "loss": 1.0036, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.03029286358765224, | |
| "grad_norm": 0.5907948017120361, | |
| "learning_rate": 4.8486936184212675e-05, | |
| "loss": 1.0408, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.030618593303648498, | |
| "grad_norm": 0.6083859801292419, | |
| "learning_rate": 4.847064916789634e-05, | |
| "loss": 1.0313, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.03094432301964476, | |
| "grad_norm": 0.5470224618911743, | |
| "learning_rate": 4.8454362151580006e-05, | |
| "loss": 1.0395, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.03127005273564102, | |
| "grad_norm": 0.9455150961875916, | |
| "learning_rate": 4.843807513526367e-05, | |
| "loss": 1.0132, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.03159578245163728, | |
| "grad_norm": 0.9068177938461304, | |
| "learning_rate": 4.8421788118947344e-05, | |
| "loss": 1.0219, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.03192151216763354, | |
| "grad_norm": 0.6018943190574646, | |
| "learning_rate": 4.840550110263101e-05, | |
| "loss": 0.9966, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.032247241883629804, | |
| "grad_norm": 1.1521615982055664, | |
| "learning_rate": 4.838921408631467e-05, | |
| "loss": 0.9782, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.03257297159962606, | |
| "grad_norm": 0.33281368017196655, | |
| "learning_rate": 4.837292706999834e-05, | |
| "loss": 1.0325, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.03289870131562232, | |
| "grad_norm": 0.8903327584266663, | |
| "learning_rate": 4.835664005368201e-05, | |
| "loss": 0.9889, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 0.03322443103161858, | |
| "grad_norm": 0.5526803731918335, | |
| "learning_rate": 4.834035303736568e-05, | |
| "loss": 1.0018, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 0.03355016074761485, | |
| "grad_norm": 0.8086706399917603, | |
| "learning_rate": 4.832406602104934e-05, | |
| "loss": 1.0189, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 0.03387589046361111, | |
| "grad_norm": 0.6990864276885986, | |
| "learning_rate": 4.830777900473301e-05, | |
| "loss": 0.996, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 0.034201620179607366, | |
| "grad_norm": 0.4859602451324463, | |
| "learning_rate": 4.829149198841668e-05, | |
| "loss": 0.992, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.034527349895603625, | |
| "grad_norm": 1.2284592390060425, | |
| "learning_rate": 4.827520497210034e-05, | |
| "loss": 1.0139, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 0.034853079611599884, | |
| "grad_norm": 0.6529733538627625, | |
| "learning_rate": 4.825891795578401e-05, | |
| "loss": 1.025, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 0.03517880932759615, | |
| "grad_norm": 0.6755232810974121, | |
| "learning_rate": 4.8242630939467674e-05, | |
| "loss": 1.0123, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 0.03550453904359241, | |
| "grad_norm": 0.9006055593490601, | |
| "learning_rate": 4.8226343923151347e-05, | |
| "loss": 0.9936, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 0.03583026875958867, | |
| "grad_norm": 0.7058572769165039, | |
| "learning_rate": 4.821005690683501e-05, | |
| "loss": 0.934, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.03615599847558493, | |
| "grad_norm": 0.4535008668899536, | |
| "learning_rate": 4.819376989051868e-05, | |
| "loss": 1.0269, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 0.036481728191581186, | |
| "grad_norm": 0.39823395013809204, | |
| "learning_rate": 4.8177482874202344e-05, | |
| "loss": 0.9866, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 0.03680745790757745, | |
| "grad_norm": 0.8109054565429688, | |
| "learning_rate": 4.816119585788601e-05, | |
| "loss": 1.0209, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 0.03713318762357371, | |
| "grad_norm": 0.760396420955658, | |
| "learning_rate": 4.814490884156968e-05, | |
| "loss": 0.9711, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 0.03745891733956997, | |
| "grad_norm": 0.8584955334663391, | |
| "learning_rate": 4.812862182525335e-05, | |
| "loss": 1.0151, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.03778464705556623, | |
| "grad_norm": 1.104041576385498, | |
| "learning_rate": 4.8112334808937013e-05, | |
| "loss": 0.9826, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 0.038110376771562496, | |
| "grad_norm": 0.6111257672309875, | |
| "learning_rate": 4.809604779262068e-05, | |
| "loss": 0.9524, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 0.038436106487558755, | |
| "grad_norm": 0.6601366996765137, | |
| "learning_rate": 4.807976077630435e-05, | |
| "loss": 0.9527, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 0.038761836203555014, | |
| "grad_norm": 0.4624398350715637, | |
| "learning_rate": 4.806347375998802e-05, | |
| "loss": 1.0077, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 0.03908756591955127, | |
| "grad_norm": 0.2786065638065338, | |
| "learning_rate": 4.8047186743671676e-05, | |
| "loss": 0.956, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.03941329563554753, | |
| "grad_norm": 1.0275955200195312, | |
| "learning_rate": 4.803089972735535e-05, | |
| "loss": 0.9484, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 0.0397390253515438, | |
| "grad_norm": 0.6198407411575317, | |
| "learning_rate": 4.8014612711039015e-05, | |
| "loss": 0.9847, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 0.04006475506754006, | |
| "grad_norm": 0.5880489945411682, | |
| "learning_rate": 4.799832569472269e-05, | |
| "loss": 0.9559, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 0.040390484783536316, | |
| "grad_norm": 0.39753594994544983, | |
| "learning_rate": 4.7982038678406346e-05, | |
| "loss": 0.9489, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 0.040716214499532576, | |
| "grad_norm": 0.5815085768699646, | |
| "learning_rate": 4.796575166209002e-05, | |
| "loss": 0.9567, | |
| "step": 12500 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 307003, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.857451720704e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |