| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9986282578875172, | |
| "eval_steps": 500, | |
| "global_step": 273, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003657978966620942, | |
| "grad_norm": 0.11399450898170471, | |
| "learning_rate": 3.5714285714285714e-06, | |
| "loss": 1.3486, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007315957933241884, | |
| "grad_norm": 0.11220081150531769, | |
| "learning_rate": 7.142857142857143e-06, | |
| "loss": 1.359, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010973936899862825, | |
| "grad_norm": 0.11029545217752457, | |
| "learning_rate": 1.0714285714285714e-05, | |
| "loss": 1.3603, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.014631915866483767, | |
| "grad_norm": 0.09306512773036957, | |
| "learning_rate": 1.4285714285714285e-05, | |
| "loss": 1.352, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01828989483310471, | |
| "grad_norm": 0.23787762224674225, | |
| "learning_rate": 1.785714285714286e-05, | |
| "loss": 1.3645, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02194787379972565, | |
| "grad_norm": 0.07387927174568176, | |
| "learning_rate": 2.1428571428571428e-05, | |
| "loss": 1.3406, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.025605852766346592, | |
| "grad_norm": 0.08182856440544128, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.3445, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.029263831732967534, | |
| "grad_norm": 0.07252974063158035, | |
| "learning_rate": 2.857142857142857e-05, | |
| "loss": 1.3225, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03292181069958848, | |
| "grad_norm": 0.07590305060148239, | |
| "learning_rate": 3.2142857142857144e-05, | |
| "loss": 1.3063, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03657978966620942, | |
| "grad_norm": 0.10524249821901321, | |
| "learning_rate": 3.571428571428572e-05, | |
| "loss": 1.3201, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.040237768632830365, | |
| "grad_norm": 0.04849069193005562, | |
| "learning_rate": 3.928571428571429e-05, | |
| "loss": 1.324, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0438957475994513, | |
| "grad_norm": 0.08398816734552383, | |
| "learning_rate": 4.2857142857142856e-05, | |
| "loss": 1.324, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04755372656607224, | |
| "grad_norm": 0.06089089438319206, | |
| "learning_rate": 4.642857142857143e-05, | |
| "loss": 1.3072, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.051211705532693184, | |
| "grad_norm": 0.07415427267551422, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3005, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.05486968449931413, | |
| "grad_norm": 0.07027556747198105, | |
| "learning_rate": 4.999816090185888e-05, | |
| "loss": 1.3036, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05852766346593507, | |
| "grad_norm": 0.05990540236234665, | |
| "learning_rate": 4.9992643878018054e-05, | |
| "loss": 1.3188, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06218564243255601, | |
| "grad_norm": 0.07783979922533035, | |
| "learning_rate": 4.99834497401854e-05, | |
| "loss": 1.2834, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06584362139917696, | |
| "grad_norm": 0.05285879224538803, | |
| "learning_rate": 4.997057984107466e-05, | |
| "loss": 1.287, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0695016003657979, | |
| "grad_norm": 0.06284961104393005, | |
| "learning_rate": 4.995403607420644e-05, | |
| "loss": 1.2819, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.07315957933241884, | |
| "grad_norm": 0.0440058633685112, | |
| "learning_rate": 4.993382087362959e-05, | |
| "loss": 1.2616, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07681755829903979, | |
| "grad_norm": 0.060883328318595886, | |
| "learning_rate": 4.9909937213563165e-05, | |
| "loss": 1.276, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08047553726566073, | |
| "grad_norm": 0.042049508541822433, | |
| "learning_rate": 4.988238860795873e-05, | |
| "loss": 1.2654, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.08413351623228166, | |
| "grad_norm": 0.07169731706380844, | |
| "learning_rate": 4.985117910998345e-05, | |
| "loss": 1.2802, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0877914951989026, | |
| "grad_norm": 0.04764694347977638, | |
| "learning_rate": 4.981631331142368e-05, | |
| "loss": 1.267, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.09144947416552354, | |
| "grad_norm": 0.05187158286571503, | |
| "learning_rate": 4.977779634200947e-05, | |
| "loss": 1.2659, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09510745313214448, | |
| "grad_norm": 0.0522884801030159, | |
| "learning_rate": 4.973563386865974e-05, | |
| "loss": 1.2462, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09876543209876543, | |
| "grad_norm": 0.03765333071351051, | |
| "learning_rate": 4.968983209464863e-05, | |
| "loss": 1.2546, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.10242341106538637, | |
| "grad_norm": 0.05279753729701042, | |
| "learning_rate": 4.9640397758692715e-05, | |
| "loss": 1.2446, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.10608139003200731, | |
| "grad_norm": 0.038350097835063934, | |
| "learning_rate": 4.9587338133959625e-05, | |
| "loss": 1.2615, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10973936899862825, | |
| "grad_norm": 0.05348701775074005, | |
| "learning_rate": 4.953066102699795e-05, | |
| "loss": 1.275, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1133973479652492, | |
| "grad_norm": 0.0379619225859642, | |
| "learning_rate": 4.947037477658865e-05, | |
| "loss": 1.2436, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.11705532693187014, | |
| "grad_norm": 0.0401979424059391, | |
| "learning_rate": 4.9406488252518203e-05, | |
| "loss": 1.2502, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.12071330589849108, | |
| "grad_norm": 0.035914115607738495, | |
| "learning_rate": 4.933901085427362e-05, | |
| "loss": 1.2489, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.12437128486511202, | |
| "grad_norm": 0.040159013122320175, | |
| "learning_rate": 4.926795250965952e-05, | |
| "loss": 1.233, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.12802926383173296, | |
| "grad_norm": 0.035101838409900665, | |
| "learning_rate": 4.9193323673337476e-05, | |
| "loss": 1.2434, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13168724279835392, | |
| "grad_norm": 0.03814801946282387, | |
| "learning_rate": 4.91151353252878e-05, | |
| "loss": 1.2406, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.13534522176497485, | |
| "grad_norm": 0.03351541608572006, | |
| "learning_rate": 4.9033398969194145e-05, | |
| "loss": 1.2213, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1390032007315958, | |
| "grad_norm": 0.036357831209897995, | |
| "learning_rate": 4.894812663075095e-05, | |
| "loss": 1.23, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.14266117969821673, | |
| "grad_norm": 0.03457951918244362, | |
| "learning_rate": 4.8859330855894156e-05, | |
| "loss": 1.2455, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1463191586648377, | |
| "grad_norm": 0.04117417335510254, | |
| "learning_rate": 4.876702470895531e-05, | |
| "loss": 1.2272, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14997713763145862, | |
| "grad_norm": 0.03363555297255516, | |
| "learning_rate": 4.867122177073948e-05, | |
| "loss": 1.2272, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.15363511659807957, | |
| "grad_norm": 0.04657221958041191, | |
| "learning_rate": 4.857193613652711e-05, | |
| "loss": 1.2417, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1572930955647005, | |
| "grad_norm": 0.03347886726260185, | |
| "learning_rate": 4.846918241400022e-05, | |
| "loss": 1.2409, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.16095107453132146, | |
| "grad_norm": 0.04122474789619446, | |
| "learning_rate": 4.836297572109323e-05, | |
| "loss": 1.2237, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1646090534979424, | |
| "grad_norm": 0.03406165540218353, | |
| "learning_rate": 4.825333168376864e-05, | |
| "loss": 1.2558, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16826703246456332, | |
| "grad_norm": 0.037185415625572205, | |
| "learning_rate": 4.8140266433718095e-05, | |
| "loss": 1.2149, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.17192501143118427, | |
| "grad_norm": 0.02867889776825905, | |
| "learning_rate": 4.8023796605988866e-05, | |
| "loss": 1.1882, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1755829903978052, | |
| "grad_norm": 0.03610045835375786, | |
| "learning_rate": 4.790393933653646e-05, | |
| "loss": 1.2114, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.17924096936442616, | |
| "grad_norm": 0.034297309815883636, | |
| "learning_rate": 4.77807122597034e-05, | |
| "loss": 1.1857, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.18289894833104708, | |
| "grad_norm": 0.03260922431945801, | |
| "learning_rate": 4.7654133505624695e-05, | |
| "loss": 1.2132, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18655692729766804, | |
| "grad_norm": 0.03398128226399422, | |
| "learning_rate": 4.752422169756048e-05, | |
| "loss": 1.2046, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.19021490626428897, | |
| "grad_norm": 0.035840485244989395, | |
| "learning_rate": 4.7390995949155915e-05, | |
| "loss": 1.2374, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.19387288523090992, | |
| "grad_norm": 0.030534490942955017, | |
| "learning_rate": 4.725447586162911e-05, | |
| "loss": 1.2076, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.03663424402475357, | |
| "learning_rate": 4.711468152088719e-05, | |
| "loss": 1.2239, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2011888431641518, | |
| "grad_norm": 0.03044678270816803, | |
| "learning_rate": 4.697163349457115e-05, | |
| "loss": 1.2117, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.20484682213077274, | |
| "grad_norm": 0.030734116211533546, | |
| "learning_rate": 4.6825352829029705e-05, | |
| "loss": 1.1859, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2085048010973937, | |
| "grad_norm": 0.03116767294704914, | |
| "learning_rate": 4.667586104622288e-05, | |
| "loss": 1.2065, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.21216278006401462, | |
| "grad_norm": 0.032509610056877136, | |
| "learning_rate": 4.652318014055547e-05, | |
| "loss": 1.2196, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.21582075903063558, | |
| "grad_norm": 0.030822426080703735, | |
| "learning_rate": 4.636733257564104e-05, | |
| "loss": 1.204, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2194787379972565, | |
| "grad_norm": 0.03399716317653656, | |
| "learning_rate": 4.620834128099696e-05, | |
| "loss": 1.1865, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22313671696387746, | |
| "grad_norm": 0.029582664370536804, | |
| "learning_rate": 4.604622964867078e-05, | |
| "loss": 1.195, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.2267946959304984, | |
| "grad_norm": 0.03362514078617096, | |
| "learning_rate": 4.588102152979863e-05, | |
| "loss": 1.2179, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.23045267489711935, | |
| "grad_norm": 0.03153723105788231, | |
| "learning_rate": 4.571274123109606e-05, | |
| "loss": 1.183, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.23411065386374028, | |
| "grad_norm": 0.03338644653558731, | |
| "learning_rate": 4.5541413511281826e-05, | |
| "loss": 1.2011, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.23776863283036123, | |
| "grad_norm": 0.02905767410993576, | |
| "learning_rate": 4.536706357743522e-05, | |
| "loss": 1.2092, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.24142661179698216, | |
| "grad_norm": 0.03598478436470032, | |
| "learning_rate": 4.518971708128737e-05, | |
| "loss": 1.1903, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.24508459076360312, | |
| "grad_norm": 0.03393587842583656, | |
| "learning_rate": 4.5009400115447206e-05, | |
| "loss": 1.1801, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.24874256973022404, | |
| "grad_norm": 0.03872101381421089, | |
| "learning_rate": 4.4826139209562444e-05, | |
| "loss": 1.1868, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.252400548696845, | |
| "grad_norm": 0.03533606603741646, | |
| "learning_rate": 4.463996132641641e-05, | |
| "loss": 1.1879, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.25605852766346593, | |
| "grad_norm": 0.03268212825059891, | |
| "learning_rate": 4.445089385796099e-05, | |
| "loss": 1.1986, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2597165066300869, | |
| "grad_norm": 0.03353170305490494, | |
| "learning_rate": 4.4258964621286584e-05, | |
| "loss": 1.1835, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.26337448559670784, | |
| "grad_norm": 0.03047579899430275, | |
| "learning_rate": 4.4064201854529365e-05, | |
| "loss": 1.1773, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.26703246456332874, | |
| "grad_norm": 0.034304771572351456, | |
| "learning_rate": 4.386663421271674e-05, | |
| "loss": 1.1814, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2706904435299497, | |
| "grad_norm": 0.03483429551124573, | |
| "learning_rate": 4.366629076355131e-05, | |
| "loss": 1.1928, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.27434842249657065, | |
| "grad_norm": 0.027306433767080307, | |
| "learning_rate": 4.346320098313429e-05, | |
| "loss": 1.1426, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2780064014631916, | |
| "grad_norm": 0.03144339472055435, | |
| "learning_rate": 4.3257394751628686e-05, | |
| "loss": 1.1639, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.2816643804298125, | |
| "grad_norm": 0.033143457025289536, | |
| "learning_rate": 4.3048902348863116e-05, | |
| "loss": 1.1879, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.28532235939643347, | |
| "grad_norm": 0.032496023923158646, | |
| "learning_rate": 4.283775444987681e-05, | |
| "loss": 1.218, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2889803383630544, | |
| "grad_norm": 0.03808918222784996, | |
| "learning_rate": 4.262398212040646e-05, | |
| "loss": 1.1989, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2926383173296754, | |
| "grad_norm": 0.036832649260759354, | |
| "learning_rate": 4.2407616812315554e-05, | |
| "loss": 1.1594, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.02746155858039856, | |
| "learning_rate": 4.218869035896697e-05, | |
| "loss": 1.1876, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.29995427526291724, | |
| "grad_norm": 0.04084876552224159, | |
| "learning_rate": 4.1967234970539384e-05, | |
| "loss": 1.1746, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.3036122542295382, | |
| "grad_norm": 0.030713148415088654, | |
| "learning_rate": 4.1743283229288245e-05, | |
| "loss": 1.179, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.30727023319615915, | |
| "grad_norm": 0.034150052815675735, | |
| "learning_rate": 4.151686808475204e-05, | |
| "loss": 1.1407, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.31092821216278005, | |
| "grad_norm": 0.035240545868873596, | |
| "learning_rate": 4.1288022848904485e-05, | |
| "loss": 1.1567, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.314586191129401, | |
| "grad_norm": 0.03663535416126251, | |
| "learning_rate": 4.10567811912534e-05, | |
| "loss": 1.1337, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.31824417009602196, | |
| "grad_norm": 0.03188696503639221, | |
| "learning_rate": 4.082317713388702e-05, | |
| "loss": 1.1751, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3219021490626429, | |
| "grad_norm": 0.03437764570116997, | |
| "learning_rate": 4.058724504646834e-05, | |
| "loss": 1.1659, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3255601280292638, | |
| "grad_norm": 0.036784157156944275, | |
| "learning_rate": 4.0349019641178443e-05, | |
| "loss": 1.1451, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.3292181069958848, | |
| "grad_norm": 0.03242890164256096, | |
| "learning_rate": 4.010853596760933e-05, | |
| "loss": 1.1479, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.33287608596250573, | |
| "grad_norm": 0.030873438343405724, | |
| "learning_rate": 3.986582940760717e-05, | |
| "loss": 1.1453, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.33653406492912663, | |
| "grad_norm": 0.036423344165086746, | |
| "learning_rate": 3.962093567006662e-05, | |
| "loss": 1.1505, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.3401920438957476, | |
| "grad_norm": 0.029906025156378746, | |
| "learning_rate": 3.937389078567708e-05, | |
| "loss": 1.1718, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.34385002286236854, | |
| "grad_norm": 0.03078794851899147, | |
| "learning_rate": 3.912473110162156e-05, | |
| "loss": 1.1562, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.3475080018289895, | |
| "grad_norm": 0.029547788202762604, | |
| "learning_rate": 3.887349327622901e-05, | |
| "loss": 1.1549, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3511659807956104, | |
| "grad_norm": 0.03439987823367119, | |
| "learning_rate": 3.8620214273580846e-05, | |
| "loss": 1.1736, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.35482395976223136, | |
| "grad_norm": 0.03436443582177162, | |
| "learning_rate": 3.83649313580725e-05, | |
| "loss": 1.1466, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.3584819387288523, | |
| "grad_norm": 0.03227129206061363, | |
| "learning_rate": 3.8107682088930794e-05, | |
| "loss": 1.1491, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.36213991769547327, | |
| "grad_norm": 0.029436925426125526, | |
| "learning_rate": 3.784850431468795e-05, | |
| "loss": 1.1497, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.36579789666209417, | |
| "grad_norm": 0.030241703614592552, | |
| "learning_rate": 3.7587436167613e-05, | |
| "loss": 1.1434, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3694558756287151, | |
| "grad_norm": 0.030009973794221878, | |
| "learning_rate": 3.7324516058101454e-05, | |
| "loss": 1.129, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3731138545953361, | |
| "grad_norm": 0.026958679780364037, | |
| "learning_rate": 3.705978266902409e-05, | |
| "loss": 1.15, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.37677183356195704, | |
| "grad_norm": 0.02721358649432659, | |
| "learning_rate": 3.679327495003561e-05, | |
| "loss": 1.137, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.38042981252857794, | |
| "grad_norm": 0.02925017662346363, | |
| "learning_rate": 3.6525032111844054e-05, | |
| "loss": 1.1456, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3840877914951989, | |
| "grad_norm": 0.026657432317733765, | |
| "learning_rate": 3.6255093620441834e-05, | |
| "loss": 1.1339, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.38774577046181985, | |
| "grad_norm": 0.02823042869567871, | |
| "learning_rate": 3.598349919129917e-05, | |
| "loss": 1.1406, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3914037494284408, | |
| "grad_norm": 0.031808383762836456, | |
| "learning_rate": 3.571028878352084e-05, | |
| "loss": 1.1561, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 0.02964746206998825, | |
| "learning_rate": 3.5435502593967104e-05, | |
| "loss": 1.1377, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.39871970736168266, | |
| "grad_norm": 0.03088381327688694, | |
| "learning_rate": 3.515918105133958e-05, | |
| "loss": 1.1499, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4023776863283036, | |
| "grad_norm": 0.04066149517893791, | |
| "learning_rate": 3.4881364810233095e-05, | |
| "loss": 1.1453, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4060356652949246, | |
| "grad_norm": 0.03428781405091286, | |
| "learning_rate": 3.460209474515428e-05, | |
| "loss": 1.1363, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.4096936442615455, | |
| "grad_norm": 0.03334755077958107, | |
| "learning_rate": 3.432141194450772e-05, | |
| "loss": 1.1414, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.41335162322816643, | |
| "grad_norm": 0.03784303367137909, | |
| "learning_rate": 3.403935770455077e-05, | |
| "loss": 1.1555, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4170096021947874, | |
| "grad_norm": 0.030056986957788467, | |
| "learning_rate": 3.375597352331772e-05, | |
| "loss": 1.1272, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.42066758116140834, | |
| "grad_norm": 0.030737608671188354, | |
| "learning_rate": 3.347130109451422e-05, | |
| "loss": 1.1271, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.42432556012802924, | |
| "grad_norm": 0.030851833522319794, | |
| "learning_rate": 3.3185382301383064e-05, | |
| "loss": 1.1368, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.4279835390946502, | |
| "grad_norm": 0.027810659259557724, | |
| "learning_rate": 3.28982592105419e-05, | |
| "loss": 1.1339, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.43164151806127116, | |
| "grad_norm": 0.030422871932387352, | |
| "learning_rate": 3.260997406579417e-05, | |
| "loss": 1.1104, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4352994970278921, | |
| "grad_norm": 0.029860053211450577, | |
| "learning_rate": 3.232056928191376e-05, | |
| "loss": 1.1138, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.438957475994513, | |
| "grad_norm": 0.031170522794127464, | |
| "learning_rate": 3.2030087438404685e-05, | |
| "loss": 1.1492, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.44261545496113397, | |
| "grad_norm": 0.027890468016266823, | |
| "learning_rate": 3.173857127323642e-05, | |
| "loss": 1.1154, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4462734339277549, | |
| "grad_norm": 0.02984224259853363, | |
| "learning_rate": 3.144606367655595e-05, | |
| "loss": 1.1299, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.4499314128943759, | |
| "grad_norm": 0.029711229726672173, | |
| "learning_rate": 3.115260768437747e-05, | |
| "loss": 1.1251, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4535893918609968, | |
| "grad_norm": 0.02858361415565014, | |
| "learning_rate": 3.085824647225056e-05, | |
| "loss": 1.1, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.45724737082761774, | |
| "grad_norm": 0.02895357273519039, | |
| "learning_rate": 3.056302334890786e-05, | |
| "loss": 1.1142, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4609053497942387, | |
| "grad_norm": 0.03063139319419861, | |
| "learning_rate": 3.0266981749893157e-05, | |
| "loss": 1.1428, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.46456332876085965, | |
| "grad_norm": 0.0273541621863842, | |
| "learning_rate": 2.997016523117081e-05, | |
| "loss": 1.1185, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.46822130772748055, | |
| "grad_norm": 0.028322231024503708, | |
| "learning_rate": 2.9672617462717444e-05, | |
| "loss": 1.1318, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.4718792866941015, | |
| "grad_norm": 0.03172619268298149, | |
| "learning_rate": 2.9374382222096886e-05, | |
| "loss": 1.136, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.47553726566072246, | |
| "grad_norm": 0.02729085460305214, | |
| "learning_rate": 2.9075503388019272e-05, | |
| "loss": 1.1297, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4791952446273434, | |
| "grad_norm": 0.02979847602546215, | |
| "learning_rate": 2.877602493388525e-05, | |
| "loss": 1.1195, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4828532235939643, | |
| "grad_norm": 0.028623459860682487, | |
| "learning_rate": 2.8475990921316292e-05, | |
| "loss": 1.1349, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4865112025605853, | |
| "grad_norm": 0.027000540867447853, | |
| "learning_rate": 2.8175445493671972e-05, | |
| "loss": 1.1132, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.49016918152720623, | |
| "grad_norm": 0.02691473811864853, | |
| "learning_rate": 2.787443286955528e-05, | |
| "loss": 1.1083, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 0.027869021520018578, | |
| "learning_rate": 2.7572997336306812e-05, | |
| "loss": 1.1345, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4974851394604481, | |
| "grad_norm": 0.029120417311787605, | |
| "learning_rate": 2.7271183243488878e-05, | |
| "loss": 1.1413, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.501143118427069, | |
| "grad_norm": 0.02720428816974163, | |
| "learning_rate": 2.696903499636045e-05, | |
| "loss": 1.1066, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.50480109739369, | |
| "grad_norm": 0.0264853835105896, | |
| "learning_rate": 2.6666597049343882e-05, | |
| "loss": 1.1257, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.508459076360311, | |
| "grad_norm": 0.02642214484512806, | |
| "learning_rate": 2.636391389948449e-05, | |
| "loss": 1.087, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.5121170553269319, | |
| "grad_norm": 0.02971823699772358, | |
| "learning_rate": 2.606103007990371e-05, | |
| "loss": 1.1263, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5157750342935528, | |
| "grad_norm": 0.026655763387680054, | |
| "learning_rate": 2.5757990153247124e-05, | |
| "loss": 1.095, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.5194330132601738, | |
| "grad_norm": 0.03130535036325455, | |
| "learning_rate": 2.5454838705127993e-05, | |
| "loss": 1.1259, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.5230909922267947, | |
| "grad_norm": 0.02821686491370201, | |
| "learning_rate": 2.5151620337567495e-05, | |
| "loss": 1.1159, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.5267489711934157, | |
| "grad_norm": 0.026856008917093277, | |
| "learning_rate": 2.4848379662432515e-05, | |
| "loss": 1.0893, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.5304069501600366, | |
| "grad_norm": 0.026246678084135056, | |
| "learning_rate": 2.4545161294872013e-05, | |
| "loss": 1.107, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5340649291266575, | |
| "grad_norm": 0.03086758591234684, | |
| "learning_rate": 2.4242009846752885e-05, | |
| "loss": 1.096, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.5377229080932785, | |
| "grad_norm": 0.03089798055589199, | |
| "learning_rate": 2.39389699200963e-05, | |
| "loss": 1.1143, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.5413808870598994, | |
| "grad_norm": 0.02865259163081646, | |
| "learning_rate": 2.3636086100515527e-05, | |
| "loss": 1.0935, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5450388660265203, | |
| "grad_norm": 0.02822408266365528, | |
| "learning_rate": 2.3333402950656124e-05, | |
| "loss": 1.092, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5486968449931413, | |
| "grad_norm": 0.027232255786657333, | |
| "learning_rate": 2.3030965003639566e-05, | |
| "loss": 1.1075, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5523548239597622, | |
| "grad_norm": 0.030684614554047585, | |
| "learning_rate": 2.272881675651112e-05, | |
| "loss": 1.0915, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5560128029263832, | |
| "grad_norm": 0.027584511786699295, | |
| "learning_rate": 2.242700266369319e-05, | |
| "loss": 1.1051, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5596707818930041, | |
| "grad_norm": 0.027764879167079926, | |
| "learning_rate": 2.2125567130444724e-05, | |
| "loss": 1.1229, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.563328760859625, | |
| "grad_norm": 0.02984793111681938, | |
| "learning_rate": 2.182455450632803e-05, | |
| "loss": 1.0987, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.566986739826246, | |
| "grad_norm": 0.03010164201259613, | |
| "learning_rate": 2.1524009078683717e-05, | |
| "loss": 1.1299, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5706447187928669, | |
| "grad_norm": 0.026011740788817406, | |
| "learning_rate": 2.1223975066114754e-05, | |
| "loss": 1.0861, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5743026977594878, | |
| "grad_norm": 0.02605109103024006, | |
| "learning_rate": 2.0924496611980734e-05, | |
| "loss": 1.1276, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.5779606767261088, | |
| "grad_norm": 0.02799968421459198, | |
| "learning_rate": 2.062561777790312e-05, | |
| "loss": 1.1116, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5816186556927297, | |
| "grad_norm": 0.02790476195514202, | |
| "learning_rate": 2.0327382537282562e-05, | |
| "loss": 1.0978, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5852766346593508, | |
| "grad_norm": 0.028209157288074493, | |
| "learning_rate": 2.0029834768829196e-05, | |
| "loss": 1.1248, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5889346136259717, | |
| "grad_norm": 0.026728255674242973, | |
| "learning_rate": 1.973301825010685e-05, | |
| "loss": 1.0831, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.02696623094379902, | |
| "learning_rate": 1.9436976651092144e-05, | |
| "loss": 1.0861, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5962505715592136, | |
| "grad_norm": 0.025825291872024536, | |
| "learning_rate": 1.9141753527749443e-05, | |
| "loss": 1.1009, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5999085505258345, | |
| "grad_norm": 0.02682458609342575, | |
| "learning_rate": 1.8847392315622535e-05, | |
| "loss": 1.1083, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.6035665294924554, | |
| "grad_norm": 0.029443522915244102, | |
| "learning_rate": 1.8553936323444058e-05, | |
| "loss": 1.1143, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6072245084590764, | |
| "grad_norm": 0.02744133211672306, | |
| "learning_rate": 1.8261428726763584e-05, | |
| "loss": 1.0903, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.6108824874256973, | |
| "grad_norm": 0.026245735585689545, | |
| "learning_rate": 1.7969912561595317e-05, | |
| "loss": 1.0843, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.6145404663923183, | |
| "grad_norm": 0.027246467769145966, | |
| "learning_rate": 1.7679430718086243e-05, | |
| "loss": 1.091, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.6181984453589392, | |
| "grad_norm": 0.026942765340209007, | |
| "learning_rate": 1.7390025934205837e-05, | |
| "loss": 1.0844, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.6218564243255601, | |
| "grad_norm": 0.02487928792834282, | |
| "learning_rate": 1.7101740789458097e-05, | |
| "loss": 1.0941, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6255144032921811, | |
| "grad_norm": 0.027003785595297813, | |
| "learning_rate": 1.6814617698616945e-05, | |
| "loss": 1.097, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.629172382258802, | |
| "grad_norm": 0.025280645117163658, | |
| "learning_rate": 1.6528698905485784e-05, | |
| "loss": 1.0773, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.6328303612254229, | |
| "grad_norm": 0.02396760694682598, | |
| "learning_rate": 1.624402647668229e-05, | |
| "loss": 1.0755, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.6364883401920439, | |
| "grad_norm": 0.023587588220834732, | |
| "learning_rate": 1.5960642295449228e-05, | |
| "loss": 1.0741, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.6401463191586648, | |
| "grad_norm": 0.024295572191476822, | |
| "learning_rate": 1.567858805549229e-05, | |
| "loss": 1.1111, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6438042981252858, | |
| "grad_norm": 0.024151455610990524, | |
| "learning_rate": 1.539790525484573e-05, | |
| "loss": 1.0789, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.6474622770919067, | |
| "grad_norm": 0.024293672293424606, | |
| "learning_rate": 1.511863518976691e-05, | |
| "loss": 1.0671, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.6511202560585276, | |
| "grad_norm": 0.025013579055666924, | |
| "learning_rate": 1.4840818948660434e-05, | |
| "loss": 1.0899, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6547782350251486, | |
| "grad_norm": 0.02457435615360737, | |
| "learning_rate": 1.456449740603291e-05, | |
| "loss": 1.0998, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6584362139917695, | |
| "grad_norm": 0.024968327954411507, | |
| "learning_rate": 1.4289711216479156e-05, | |
| "loss": 1.0846, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6620941929583904, | |
| "grad_norm": 0.023626726120710373, | |
| "learning_rate": 1.401650080870083e-05, | |
| "loss": 1.0774, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6657521719250115, | |
| "grad_norm": 0.025500517338514328, | |
| "learning_rate": 1.3744906379558165e-05, | |
| "loss": 1.096, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6694101508916324, | |
| "grad_norm": 0.02676280215382576, | |
| "learning_rate": 1.3474967888155948e-05, | |
| "loss": 1.0984, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6730681298582533, | |
| "grad_norm": 0.024672549217939377, | |
| "learning_rate": 1.3206725049964392e-05, | |
| "loss": 1.1026, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.6767261088248743, | |
| "grad_norm": 0.02653788961470127, | |
| "learning_rate": 1.2940217330975912e-05, | |
| "loss": 1.0898, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6803840877914952, | |
| "grad_norm": 0.024248803034424782, | |
| "learning_rate": 1.2675483941898548e-05, | |
| "loss": 1.0925, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6840420667581162, | |
| "grad_norm": 0.02738628350198269, | |
| "learning_rate": 1.2412563832387003e-05, | |
| "loss": 1.0916, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.6877000457247371, | |
| "grad_norm": 0.023650605231523514, | |
| "learning_rate": 1.2151495685312054e-05, | |
| "loss": 1.0862, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.691358024691358, | |
| "grad_norm": 0.024786144495010376, | |
| "learning_rate": 1.1892317911069212e-05, | |
| "loss": 1.0791, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.695016003657979, | |
| "grad_norm": 0.026861809194087982, | |
| "learning_rate": 1.163506864192751e-05, | |
| "loss": 1.1008, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6986739826245999, | |
| "grad_norm": 0.024741502478718758, | |
| "learning_rate": 1.1379785726419162e-05, | |
| "loss": 1.0804, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.7023319615912208, | |
| "grad_norm": 0.023557139560580254, | |
| "learning_rate": 1.1126506723770996e-05, | |
| "loss": 1.0503, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.7059899405578418, | |
| "grad_norm": 0.023487713187932968, | |
| "learning_rate": 1.087526889837845e-05, | |
| "loss": 1.0778, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.7096479195244627, | |
| "grad_norm": 0.025643397122621536, | |
| "learning_rate": 1.0626109214322923e-05, | |
| "loss": 1.081, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.7133058984910837, | |
| "grad_norm": 0.02375604212284088, | |
| "learning_rate": 1.0379064329933385e-05, | |
| "loss": 1.0659, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.7169638774577046, | |
| "grad_norm": 0.02461252547800541, | |
| "learning_rate": 1.0134170592392836e-05, | |
| "loss": 1.0785, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.7206218564243255, | |
| "grad_norm": 0.024264885112643242, | |
| "learning_rate": 9.89146403239067e-06, | |
| "loss": 1.0868, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.7242798353909465, | |
| "grad_norm": 0.02400391362607479, | |
| "learning_rate": 9.650980358821559e-06, | |
| "loss": 1.0826, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.7279378143575674, | |
| "grad_norm": 0.025086689740419388, | |
| "learning_rate": 9.412754953531663e-06, | |
| "loss": 1.0726, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.7315957933241883, | |
| "grad_norm": 0.023492420092225075, | |
| "learning_rate": 9.176822866112988e-06, | |
| "loss": 1.0781, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7352537722908093, | |
| "grad_norm": 0.022942854091525078, | |
| "learning_rate": 8.943218808746603e-06, | |
| "loss": 1.0589, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.7389117512574302, | |
| "grad_norm": 0.0239090695977211, | |
| "learning_rate": 8.711977151095524e-06, | |
| "loss": 1.0627, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.7425697302240513, | |
| "grad_norm": 0.02404557354748249, | |
| "learning_rate": 8.483131915247968e-06, | |
| "loss": 1.0707, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.7462277091906722, | |
| "grad_norm": 0.02398700639605522, | |
| "learning_rate": 8.256716770711764e-06, | |
| "loss": 1.0637, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.7498856881572931, | |
| "grad_norm": 0.02359054796397686, | |
| "learning_rate": 8.032765029460625e-06, | |
| "loss": 1.0806, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7535436671239141, | |
| "grad_norm": 0.022856276482343674, | |
| "learning_rate": 7.811309641033041e-06, | |
| "loss": 1.0466, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.757201646090535, | |
| "grad_norm": 0.02421443723142147, | |
| "learning_rate": 7.592383187684457e-06, | |
| "loss": 1.0805, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7608596250571559, | |
| "grad_norm": 0.025367150083184242, | |
| "learning_rate": 7.376017879593553e-06, | |
| "loss": 1.098, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7645176040237769, | |
| "grad_norm": 0.02458193711936474, | |
| "learning_rate": 7.1622455501231925e-06, | |
| "loss": 1.0872, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.7681755829903978, | |
| "grad_norm": 0.02363410033285618, | |
| "learning_rate": 6.951097651136889e-06, | |
| "loss": 1.0892, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7718335619570188, | |
| "grad_norm": 0.023377366364002228, | |
| "learning_rate": 6.74260524837132e-06, | |
| "loss": 1.0703, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7754915409236397, | |
| "grad_norm": 0.02378440462052822, | |
| "learning_rate": 6.536799016865713e-06, | |
| "loss": 1.0825, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7791495198902606, | |
| "grad_norm": 0.023137414827942848, | |
| "learning_rate": 6.333709236448693e-06, | |
| "loss": 1.054, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7828074988568816, | |
| "grad_norm": 0.024228321388363838, | |
| "learning_rate": 6.133365787283268e-06, | |
| "loss": 1.0677, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7864654778235025, | |
| "grad_norm": 0.02322709932923317, | |
| "learning_rate": 5.935798145470639e-06, | |
| "loss": 1.0539, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.02640087902545929, | |
| "learning_rate": 5.741035378713428e-06, | |
| "loss": 1.0623, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7937814357567444, | |
| "grad_norm": 0.024857627227902412, | |
| "learning_rate": 5.549106142039018e-06, | |
| "loss": 1.0866, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.7974394147233653, | |
| "grad_norm": 0.02291812375187874, | |
| "learning_rate": 5.360038673583601e-06, | |
| "loss": 1.0728, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.8010973936899863, | |
| "grad_norm": 0.02379731461405754, | |
| "learning_rate": 5.173860790437563e-06, | |
| "loss": 1.0333, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.8047553726566072, | |
| "grad_norm": 0.02438983879983425, | |
| "learning_rate": 4.990599884552802e-06, | |
| "loss": 1.0556, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8084133516232281, | |
| "grad_norm": 0.023418473079800606, | |
| "learning_rate": 4.8102829187126355e-06, | |
| "loss": 1.0908, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.8120713305898491, | |
| "grad_norm": 0.022404542192816734, | |
| "learning_rate": 4.632936422564785e-06, | |
| "loss": 1.0702, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.81572930955647, | |
| "grad_norm": 0.024710949510335922, | |
| "learning_rate": 4.458586488718178e-06, | |
| "loss": 1.0651, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.819387288523091, | |
| "grad_norm": 0.023974299430847168, | |
| "learning_rate": 4.2872587689039484e-06, | |
| "loss": 1.0798, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.823045267489712, | |
| "grad_norm": 0.022725731134414673, | |
| "learning_rate": 4.118978470201379e-06, | |
| "loss": 1.0739, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8267032464563329, | |
| "grad_norm": 0.023963825777173042, | |
| "learning_rate": 3.953770351329228e-06, | |
| "loss": 1.0608, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.8303612254229538, | |
| "grad_norm": 0.02529250457882881, | |
| "learning_rate": 3.7916587190030467e-06, | |
| "loss": 1.0827, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.8340192043895748, | |
| "grad_norm": 0.023284699767827988, | |
| "learning_rate": 3.6326674243589653e-06, | |
| "loss": 1.0715, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.8376771833561957, | |
| "grad_norm": 0.02325253002345562, | |
| "learning_rate": 3.476819859444538e-06, | |
| "loss": 1.068, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.8413351623228167, | |
| "grad_norm": 0.02358105778694153, | |
| "learning_rate": 3.3241389537771174e-06, | |
| "loss": 1.0859, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.8449931412894376, | |
| "grad_norm": 0.023099737241864204, | |
| "learning_rate": 3.1746471709702964e-06, | |
| "loss": 1.0818, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.8486511202560585, | |
| "grad_norm": 0.02324778586626053, | |
| "learning_rate": 3.028366505428856e-06, | |
| "loss": 1.0371, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.8523090992226795, | |
| "grad_norm": 0.02218066342175007, | |
| "learning_rate": 2.8853184791128086e-06, | |
| "loss": 1.1001, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.8559670781893004, | |
| "grad_norm": 0.022952284663915634, | |
| "learning_rate": 2.745524138370892e-06, | |
| "loss": 1.0495, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.8596250571559213, | |
| "grad_norm": 0.022721242159605026, | |
| "learning_rate": 2.6090040508440857e-06, | |
| "loss": 1.0614, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8632830361225423, | |
| "grad_norm": 0.02230103127658367, | |
| "learning_rate": 2.475778302439524e-06, | |
| "loss": 1.0881, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.8669410150891632, | |
| "grad_norm": 0.023484766483306885, | |
| "learning_rate": 2.345866494375307e-06, | |
| "loss": 1.0753, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.8705989940557842, | |
| "grad_norm": 0.024059921503067017, | |
| "learning_rate": 2.219287740296605e-06, | |
| "loss": 1.0512, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8742569730224051, | |
| "grad_norm": 0.02326652780175209, | |
| "learning_rate": 2.0960606634635365e-06, | |
| "loss": 1.085, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.877914951989026, | |
| "grad_norm": 0.024198243394494057, | |
| "learning_rate": 1.9762033940111335e-06, | |
| "loss": 1.0602, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.881572930955647, | |
| "grad_norm": 0.02250683680176735, | |
| "learning_rate": 1.8597335662819099e-06, | |
| "loss": 1.0696, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8852309099222679, | |
| "grad_norm": 0.02158784680068493, | |
| "learning_rate": 1.746668316231359e-06, | |
| "loss": 1.0706, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.023256294429302216, | |
| "learning_rate": 1.6370242789067764e-06, | |
| "loss": 1.103, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8925468678555099, | |
| "grad_norm": 0.02234894596040249, | |
| "learning_rate": 1.530817585999783e-06, | |
| "loss": 1.0722, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8962048468221308, | |
| "grad_norm": 0.02256467379629612, | |
| "learning_rate": 1.428063863472895e-06, | |
| "loss": 1.0671, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8998628257887518, | |
| "grad_norm": 0.02275981567800045, | |
| "learning_rate": 1.3287782292605245e-06, | |
| "loss": 1.0957, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.9035208047553727, | |
| "grad_norm": 0.022567572072148323, | |
| "learning_rate": 1.2329752910446913e-06, | |
| "loss": 1.0343, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.9071787837219936, | |
| "grad_norm": 0.022027762606739998, | |
| "learning_rate": 1.1406691441058493e-06, | |
| "loss": 1.0584, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.9108367626886146, | |
| "grad_norm": 0.022433646023273468, | |
| "learning_rate": 1.0518733692490513e-06, | |
| "loss": 1.0421, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.9144947416552355, | |
| "grad_norm": 0.023060685023665428, | |
| "learning_rate": 9.66601030805861e-07, | |
| "loss": 1.057, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9181527206218564, | |
| "grad_norm": 0.021818097680807114, | |
| "learning_rate": 8.848646747122075e-07, | |
| "loss": 1.0751, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.9218106995884774, | |
| "grad_norm": 0.02274404466152191, | |
| "learning_rate": 8.066763266625282e-07, | |
| "loss": 1.0533, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.9254686785550983, | |
| "grad_norm": 0.02207763120532036, | |
| "learning_rate": 7.320474903404762e-07, | |
| "loss": 1.0718, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.9291266575217193, | |
| "grad_norm": 0.021938692778348923, | |
| "learning_rate": 6.609891457263817e-07, | |
| "loss": 1.0668, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.9327846364883402, | |
| "grad_norm": 0.023405423387885094, | |
| "learning_rate": 5.935117474818036e-07, | |
| "loss": 1.0688, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9364426154549611, | |
| "grad_norm": 0.02231016755104065, | |
| "learning_rate": 5.296252234113564e-07, | |
| "loss": 1.0612, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.9401005944215821, | |
| "grad_norm": 0.022039178758859634, | |
| "learning_rate": 4.6933897300205085e-07, | |
| "loss": 1.0607, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.943758573388203, | |
| "grad_norm": 0.02199982851743698, | |
| "learning_rate": 4.1266186604037307e-07, | |
| "loss": 1.0678, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.9474165523548239, | |
| "grad_norm": 0.021750394254922867, | |
| "learning_rate": 3.5960224130728857e-07, | |
| "loss": 1.0511, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.9510745313214449, | |
| "grad_norm": 0.021463261917233467, | |
| "learning_rate": 3.101679053513745e-07, | |
| "loss": 1.0586, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9547325102880658, | |
| "grad_norm": 0.022153394296765327, | |
| "learning_rate": 2.643661313402601e-07, | |
| "loss": 1.0655, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.9583904892546868, | |
| "grad_norm": 0.022121533751487732, | |
| "learning_rate": 2.2220365799053843e-07, | |
| "loss": 1.0519, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.9620484682213077, | |
| "grad_norm": 0.022767378017306328, | |
| "learning_rate": 1.836866885763233e-07, | |
| "loss": 1.097, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.9657064471879286, | |
| "grad_norm": 0.021641606464982033, | |
| "learning_rate": 1.488208900165572e-07, | |
| "loss": 1.0564, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.9693644261545497, | |
| "grad_norm": 0.022372765466570854, | |
| "learning_rate": 1.176113920412697e-07, | |
| "loss": 1.0683, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9730224051211706, | |
| "grad_norm": 0.021753614768385887, | |
| "learning_rate": 9.006278643683696e-08, | |
| "loss": 1.0648, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.9766803840877915, | |
| "grad_norm": 0.022067097947001457, | |
| "learning_rate": 6.617912637040812e-08, | |
| "loss": 1.0781, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9803383630544125, | |
| "grad_norm": 0.021563511341810226, | |
| "learning_rate": 4.596392579356845e-08, | |
| "loss": 1.0615, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9839963420210334, | |
| "grad_norm": 0.022357366979122162, | |
| "learning_rate": 2.9420158925341777e-08, | |
| "loss": 1.0623, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.02167103812098503, | |
| "learning_rate": 1.6550259814601676e-08, | |
| "loss": 1.0464, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9913122999542753, | |
| "grad_norm": 0.021607818081974983, | |
| "learning_rate": 7.356121981946795e-09, | |
| "loss": 1.0607, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9949702789208962, | |
| "grad_norm": 0.022355254739522934, | |
| "learning_rate": 1.8390981411264829e-09, | |
| "loss": 1.0465, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.9986282578875172, | |
| "grad_norm": 0.022043729200959206, | |
| "learning_rate": 0.0, | |
| "loss": 1.0636, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9986282578875172, | |
| "step": 273, | |
| "total_flos": 1921476294868992.0, | |
| "train_loss": 1.1405032803287436, | |
| "train_runtime": 59872.6859, | |
| "train_samples_per_second": 0.584, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 273, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1921476294868992.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |