| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.477611940298507, | |
| "eval_steps": 500, | |
| "global_step": 1200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018656716417910446, | |
| "grad_norm": 2.403158187866211, | |
| "learning_rate": 4.9998282347929784e-05, | |
| "loss": 3.3875, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03731343283582089, | |
| "grad_norm": 2.301710367202759, | |
| "learning_rate": 4.99931296277454e-05, | |
| "loss": 2.9015, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.055970149253731345, | |
| "grad_norm": 1.271048665046692, | |
| "learning_rate": 4.998454254749331e-05, | |
| "loss": 2.6229, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 1.069893717765808, | |
| "learning_rate": 4.997252228714279e-05, | |
| "loss": 2.3704, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09328358208955224, | |
| "grad_norm": 0.9044906497001648, | |
| "learning_rate": 4.9957070498423854e-05, | |
| "loss": 2.3782, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.11194029850746269, | |
| "grad_norm": 0.9635376334190369, | |
| "learning_rate": 4.993818930460026e-05, | |
| "loss": 2.3576, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13059701492537312, | |
| "grad_norm": 0.8513979315757751, | |
| "learning_rate": 4.9915881300177725e-05, | |
| "loss": 2.4603, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 0.845267117023468, | |
| "learning_rate": 4.9890149550547454e-05, | |
| "loss": 2.2033, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16791044776119404, | |
| "grad_norm": 0.6632418036460876, | |
| "learning_rate": 4.98609975915649e-05, | |
| "loss": 2.1851, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1865671641791045, | |
| "grad_norm": 0.6857479810714722, | |
| "learning_rate": 4.982842942906386e-05, | |
| "loss": 2.3592, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.20522388059701493, | |
| "grad_norm": 0.7204287648200989, | |
| "learning_rate": 4.979244953830608e-05, | |
| "loss": 2.1323, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 0.6864420175552368, | |
| "learning_rate": 4.9753062863366276e-05, | |
| "loss": 2.2138, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.24253731343283583, | |
| "grad_norm": 0.7536088228225708, | |
| "learning_rate": 4.971027481645274e-05, | |
| "loss": 2.2584, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.26119402985074625, | |
| "grad_norm": 0.9708526134490967, | |
| "learning_rate": 4.966409127716367e-05, | |
| "loss": 2.2669, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2798507462686567, | |
| "grad_norm": 0.7516190409660339, | |
| "learning_rate": 4.96145185916792e-05, | |
| "loss": 2.2133, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 0.7864778637886047, | |
| "learning_rate": 4.95615635718894e-05, | |
| "loss": 2.1683, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.31716417910447764, | |
| "grad_norm": 0.7846741080284119, | |
| "learning_rate": 4.950523349445824e-05, | |
| "loss": 2.1274, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3358208955223881, | |
| "grad_norm": 0.816838800907135, | |
| "learning_rate": 4.944553609982363e-05, | |
| "loss": 2.2033, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.35447761194029853, | |
| "grad_norm": 0.7661916017532349, | |
| "learning_rate": 4.938247959113386e-05, | |
| "loss": 2.1492, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 0.8964986205101013, | |
| "learning_rate": 4.931607263312032e-05, | |
| "loss": 2.0862, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3917910447761194, | |
| "grad_norm": 0.8603547215461731, | |
| "learning_rate": 4.924632435090696e-05, | |
| "loss": 2.1444, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.41044776119402987, | |
| "grad_norm": 0.8611045479774475, | |
| "learning_rate": 4.917324432875627e-05, | |
| "loss": 2.1202, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4291044776119403, | |
| "grad_norm": 0.9499636888504028, | |
| "learning_rate": 4.909684260875235e-05, | |
| "loss": 2.1285, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 0.8490393161773682, | |
| "learning_rate": 4.9017129689421e-05, | |
| "loss": 2.236, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4664179104477612, | |
| "grad_norm": 0.9628555178642273, | |
| "learning_rate": 4.893411652428712e-05, | |
| "loss": 2.1219, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.48507462686567165, | |
| "grad_norm": 1.1119599342346191, | |
| "learning_rate": 4.8847814520369475e-05, | |
| "loss": 2.2537, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.503731343283582, | |
| "grad_norm": 0.9489665627479553, | |
| "learning_rate": 4.875823553661334e-05, | |
| "loss": 2.1018, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 0.9434083700180054, | |
| "learning_rate": 4.8665391882260856e-05, | |
| "loss": 2.0809, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.5410447761194029, | |
| "grad_norm": 0.8856557607650757, | |
| "learning_rate": 4.856929631515964e-05, | |
| "loss": 2.0807, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5597014925373134, | |
| "grad_norm": 0.8770031929016113, | |
| "learning_rate": 4.846996204000967e-05, | |
| "loss": 2.0843, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5783582089552238, | |
| "grad_norm": 0.8374930620193481, | |
| "learning_rate": 4.8367402706548805e-05, | |
| "loss": 2.1869, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 1.0829132795333862, | |
| "learning_rate": 4.8261632407677174e-05, | |
| "loss": 2.028, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.6156716417910447, | |
| "grad_norm": 0.9735206365585327, | |
| "learning_rate": 4.815266567752059e-05, | |
| "loss": 2.0966, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.6343283582089553, | |
| "grad_norm": 1.087944746017456, | |
| "learning_rate": 4.804051748943343e-05, | |
| "loss": 2.0863, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.6529850746268657, | |
| "grad_norm": 0.8176729083061218, | |
| "learning_rate": 4.792520325394111e-05, | |
| "loss": 2.1135, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 0.9173070788383484, | |
| "learning_rate": 4.780673881662242e-05, | |
| "loss": 2.0564, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6902985074626866, | |
| "grad_norm": 0.9463202953338623, | |
| "learning_rate": 4.7685140455932267e-05, | |
| "loss": 2.1579, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.7089552238805971, | |
| "grad_norm": 1.149950385093689, | |
| "learning_rate": 4.756042488096471e-05, | |
| "loss": 2.1447, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.7276119402985075, | |
| "grad_norm": 0.940965473651886, | |
| "learning_rate": 4.743260922915701e-05, | |
| "loss": 2.0823, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 0.9384671449661255, | |
| "learning_rate": 4.730171106393466e-05, | |
| "loss": 2.1445, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7649253731343284, | |
| "grad_norm": 0.8937250971794128, | |
| "learning_rate": 4.716774837229804e-05, | |
| "loss": 2.014, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7835820895522388, | |
| "grad_norm": 0.8928058743476868, | |
| "learning_rate": 4.7030739562350713e-05, | |
| "loss": 2.1882, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8022388059701493, | |
| "grad_norm": 1.0239906311035156, | |
| "learning_rate": 4.6890703460769955e-05, | |
| "loss": 2.1042, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 1.0555064678192139, | |
| "learning_rate": 4.674765931021976e-05, | |
| "loss": 2.015, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.8395522388059702, | |
| "grad_norm": 1.084709882736206, | |
| "learning_rate": 4.6601626766706626e-05, | |
| "loss": 2.0603, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8582089552238806, | |
| "grad_norm": 0.9265861511230469, | |
| "learning_rate": 4.645262589687861e-05, | |
| "loss": 2.1006, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.8768656716417911, | |
| "grad_norm": 1.0058296918869019, | |
| "learning_rate": 4.6300677175267914e-05, | |
| "loss": 2.063, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 1.0766576528549194, | |
| "learning_rate": 4.614580148147744e-05, | |
| "loss": 2.0781, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.914179104477612, | |
| "grad_norm": 1.0215730667114258, | |
| "learning_rate": 4.598802009731167e-05, | |
| "loss": 2.1774, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.9328358208955224, | |
| "grad_norm": 0.9870419502258301, | |
| "learning_rate": 4.582735470385229e-05, | |
| "loss": 1.9636, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.9514925373134329, | |
| "grad_norm": 1.1921675205230713, | |
| "learning_rate": 4.5663827378478975e-05, | |
| "loss": 2.0141, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 1.0618964433670044, | |
| "learning_rate": 4.5497460591835615e-05, | |
| "loss": 2.0508, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.9888059701492538, | |
| "grad_norm": 0.9723111391067505, | |
| "learning_rate": 4.532827720474268e-05, | |
| "loss": 2.0312, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.007462686567164, | |
| "grad_norm": 0.9339023232460022, | |
| "learning_rate": 4.515630046505575e-05, | |
| "loss": 2.1107, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0261194029850746, | |
| "grad_norm": 1.0588074922561646, | |
| "learning_rate": 4.498155400447107e-05, | |
| "loss": 2.0963, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 1.0709750652313232, | |
| "learning_rate": 4.480406183527823e-05, | |
| "loss": 2.0359, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.0634328358208955, | |
| "grad_norm": 1.2172249555587769, | |
| "learning_rate": 4.462384834706058e-05, | |
| "loss": 2.1083, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.0820895522388059, | |
| "grad_norm": 1.1719626188278198, | |
| "learning_rate": 4.4440938303343804e-05, | |
| "loss": 2.1259, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1007462686567164, | |
| "grad_norm": 1.051269292831421, | |
| "learning_rate": 4.425535683819312e-05, | |
| "loss": 2.0901, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 1.3167760372161865, | |
| "learning_rate": 4.406712945275955e-05, | |
| "loss": 2.0032, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.1380597014925373, | |
| "grad_norm": 1.2565367221832275, | |
| "learning_rate": 4.387628201177577e-05, | |
| "loss": 2.0148, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.1567164179104479, | |
| "grad_norm": 1.1141688823699951, | |
| "learning_rate": 4.368284074000193e-05, | |
| "loss": 2.0217, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.1753731343283582, | |
| "grad_norm": 1.1642612218856812, | |
| "learning_rate": 4.348683221862212e-05, | |
| "loss": 2.0194, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 1.1613104343414307, | |
| "learning_rate": 4.328828338159173e-05, | |
| "loss": 1.9371, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.212686567164179, | |
| "grad_norm": 1.2319557666778564, | |
| "learning_rate": 4.3087221511936434e-05, | |
| "loss": 2.0227, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.2313432835820897, | |
| "grad_norm": 1.2520420551300049, | |
| "learning_rate": 4.288367423800319e-05, | |
| "loss": 1.9883, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.0452089309692383, | |
| "learning_rate": 4.267766952966369e-05, | |
| "loss": 1.9912, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 0.9965611100196838, | |
| "learning_rate": 4.2469235694471043e-05, | |
| "loss": 1.983, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.287313432835821, | |
| "grad_norm": 1.0808607339859009, | |
| "learning_rate": 4.225840137376993e-05, | |
| "loss": 1.9514, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.3059701492537314, | |
| "grad_norm": 1.102575659751892, | |
| "learning_rate": 4.204519553876095e-05, | |
| "loss": 2.0286, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.3246268656716418, | |
| "grad_norm": 1.0246608257293701, | |
| "learning_rate": 4.1829647486519596e-05, | |
| "loss": 2.0265, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 1.0723367929458618, | |
| "learning_rate": 4.161178683597054e-05, | |
| "loss": 2.0077, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.3619402985074627, | |
| "grad_norm": 1.4298617839813232, | |
| "learning_rate": 4.139164352381758e-05, | |
| "loss": 2.0898, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.3805970149253732, | |
| "grad_norm": 1.1437115669250488, | |
| "learning_rate": 4.116924780042997e-05, | |
| "loss": 2.024, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.3992537313432836, | |
| "grad_norm": 1.326556921005249, | |
| "learning_rate": 4.094463022568569e-05, | |
| "loss": 2.2252, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 1.2549344301223755, | |
| "learning_rate": 4.071782166477213e-05, | |
| "loss": 1.9777, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.4365671641791045, | |
| "grad_norm": 1.1226497888565063, | |
| "learning_rate": 4.0488853283944806e-05, | |
| "loss": 2.0062, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.455223880597015, | |
| "grad_norm": 1.2250981330871582, | |
| "learning_rate": 4.0257756546244804e-05, | |
| "loss": 1.9147, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.4738805970149254, | |
| "grad_norm": 1.3552589416503906, | |
| "learning_rate": 4.0024563207175316e-05, | |
| "loss": 1.9709, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 1.3661599159240723, | |
| "learning_rate": 3.978930531033807e-05, | |
| "loss": 1.9748, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5111940298507462, | |
| "grad_norm": 1.1794605255126953, | |
| "learning_rate": 3.9552015183030136e-05, | |
| "loss": 2.0367, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.5298507462686568, | |
| "grad_norm": 1.19724440574646, | |
| "learning_rate": 3.93127254318018e-05, | |
| "loss": 1.9545, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.5485074626865671, | |
| "grad_norm": 1.310658574104309, | |
| "learning_rate": 3.907146893797599e-05, | |
| "loss": 1.9933, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 1.2032736539840698, | |
| "learning_rate": 3.882827885312999e-05, | |
| "loss": 2.0442, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.585820895522388, | |
| "grad_norm": 1.2670124769210815, | |
| "learning_rate": 3.858318859454001e-05, | |
| "loss": 1.974, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.6044776119402986, | |
| "grad_norm": 1.5301685333251953, | |
| "learning_rate": 3.833623184058926e-05, | |
| "loss": 2.0865, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.623134328358209, | |
| "grad_norm": 1.3863707780838013, | |
| "learning_rate": 3.808744252614012e-05, | |
| "loss": 1.9614, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 1.2591431140899658, | |
| "learning_rate": 3.783685483787105e-05, | |
| "loss": 1.949, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.6604477611940298, | |
| "grad_norm": 1.2093037366867065, | |
| "learning_rate": 3.758450320957899e-05, | |
| "loss": 1.9618, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.6791044776119404, | |
| "grad_norm": 1.1593824625015259, | |
| "learning_rate": 3.7330422317447685e-05, | |
| "loss": 2.0124, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.6977611940298507, | |
| "grad_norm": 1.7013437747955322, | |
| "learning_rate": 3.707464707528275e-05, | |
| "loss": 2.0613, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 1.2550350427627563, | |
| "learning_rate": 3.681721262971413e-05, | |
| "loss": 2.1354, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.7350746268656716, | |
| "grad_norm": 1.1735903024673462, | |
| "learning_rate": 3.6558154355366506e-05, | |
| "loss": 1.9618, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.7537313432835822, | |
| "grad_norm": 1.331148624420166, | |
| "learning_rate": 3.6297507849998344e-05, | |
| "loss": 1.9245, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.7723880597014925, | |
| "grad_norm": 1.3502494096755981, | |
| "learning_rate": 3.6035308929610446e-05, | |
| "loss": 1.9758, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 1.2406198978424072, | |
| "learning_rate": 3.5771593623524265e-05, | |
| "loss": 1.9824, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.8097014925373134, | |
| "grad_norm": 1.224885106086731, | |
| "learning_rate": 3.550639816943111e-05, | |
| "loss": 2.069, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.828358208955224, | |
| "grad_norm": 1.4666011333465576, | |
| "learning_rate": 3.5239759008412666e-05, | |
| "loss": 2.0797, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.8470149253731343, | |
| "grad_norm": 1.2758076190948486, | |
| "learning_rate": 3.497171277993346e-05, | |
| "loss": 2.0195, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 1.1991291046142578, | |
| "learning_rate": 3.4702296316806244e-05, | |
| "loss": 1.9558, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.8843283582089554, | |
| "grad_norm": 1.2548415660858154, | |
| "learning_rate": 3.443154664013067e-05, | |
| "loss": 1.9805, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.9029850746268657, | |
| "grad_norm": 1.5407222509384155, | |
| "learning_rate": 3.415950095420616e-05, | |
| "loss": 1.9152, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.921641791044776, | |
| "grad_norm": 1.285704493522644, | |
| "learning_rate": 3.3886196641419545e-05, | |
| "loss": 2.0442, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 1.377465844154358, | |
| "learning_rate": 3.361167125710832e-05, | |
| "loss": 2.0537, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.9589552238805972, | |
| "grad_norm": 1.186889410018921, | |
| "learning_rate": 3.333596252440008e-05, | |
| "loss": 1.9798, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.9776119402985075, | |
| "grad_norm": 1.4855142831802368, | |
| "learning_rate": 3.305910832902884e-05, | |
| "loss": 2.0984, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.9962686567164178, | |
| "grad_norm": 1.47159743309021, | |
| "learning_rate": 3.278114671412917e-05, | |
| "loss": 1.9932, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.014925373134328, | |
| "grad_norm": 1.4078409671783447, | |
| "learning_rate": 3.2502115875008524e-05, | |
| "loss": 1.9457, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.033582089552239, | |
| "grad_norm": 1.1386340856552124, | |
| "learning_rate": 3.222205415389877e-05, | |
| "loss": 1.9334, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.0522388059701493, | |
| "grad_norm": 1.666084885597229, | |
| "learning_rate": 3.1941000034687515e-05, | |
| "loss": 1.9716, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.0708955223880596, | |
| "grad_norm": 1.3137987852096558, | |
| "learning_rate": 3.165899213762995e-05, | |
| "loss": 1.9189, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.08955223880597, | |
| "grad_norm": 1.2372797727584839, | |
| "learning_rate": 3.1376069214041913e-05, | |
| "loss": 2.0234, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.1082089552238807, | |
| "grad_norm": 1.3149720430374146, | |
| "learning_rate": 3.109227014097505e-05, | |
| "loss": 2.0271, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.126865671641791, | |
| "grad_norm": 1.4162675142288208, | |
| "learning_rate": 3.0807633915874584e-05, | |
| "loss": 1.8236, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.1455223880597014, | |
| "grad_norm": 1.4029136896133423, | |
| "learning_rate": 3.052219965122062e-05, | |
| "loss": 2.1821, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.1641791044776117, | |
| "grad_norm": 1.5424753427505493, | |
| "learning_rate": 3.0236006569153617e-05, | |
| "loss": 1.9496, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.1828358208955225, | |
| "grad_norm": 1.274217963218689, | |
| "learning_rate": 2.9949093996084747e-05, | |
| "loss": 2.0439, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.201492537313433, | |
| "grad_norm": 1.2068248987197876, | |
| "learning_rate": 2.9661501357292033e-05, | |
| "loss": 2.0805, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.220149253731343, | |
| "grad_norm": 1.2352491617202759, | |
| "learning_rate": 2.9373268171502777e-05, | |
| "loss": 1.975, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.2388059701492535, | |
| "grad_norm": 1.3039956092834473, | |
| "learning_rate": 2.9084434045463255e-05, | |
| "loss": 1.9834, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.2574626865671643, | |
| "grad_norm": 1.3400136232376099, | |
| "learning_rate": 2.8795038668496222e-05, | |
| "loss": 1.84, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.2761194029850746, | |
| "grad_norm": 1.458132028579712, | |
| "learning_rate": 2.850512180704715e-05, | |
| "loss": 1.9518, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.294776119402985, | |
| "grad_norm": 1.446595311164856, | |
| "learning_rate": 2.821472329921981e-05, | |
| "loss": 1.8977, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.3134328358208958, | |
| "grad_norm": 1.432244062423706, | |
| "learning_rate": 2.792388304930207e-05, | |
| "loss": 1.9742, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.332089552238806, | |
| "grad_norm": 1.499017596244812, | |
| "learning_rate": 2.7632641022282502e-05, | |
| "loss": 1.9379, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.3507462686567164, | |
| "grad_norm": 1.6504281759262085, | |
| "learning_rate": 2.7341037238358774e-05, | |
| "loss": 1.9175, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.3694029850746268, | |
| "grad_norm": 1.6585911512374878, | |
| "learning_rate": 2.704911176743833e-05, | |
| "loss": 2.0449, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.388059701492537, | |
| "grad_norm": 1.545623779296875, | |
| "learning_rate": 2.6756904723632324e-05, | |
| "loss": 2.0096, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.406716417910448, | |
| "grad_norm": 1.468853235244751, | |
| "learning_rate": 2.646445625974347e-05, | |
| "loss": 1.878, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.425373134328358, | |
| "grad_norm": 1.3598605394363403, | |
| "learning_rate": 2.6171806561748502e-05, | |
| "loss": 1.9625, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.4440298507462686, | |
| "grad_norm": 1.3197077512741089, | |
| "learning_rate": 2.5878995843276204e-05, | |
| "loss": 1.9375, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.4626865671641793, | |
| "grad_norm": 1.5469880104064941, | |
| "learning_rate": 2.5586064340081516e-05, | |
| "loss": 1.8402, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.4813432835820897, | |
| "grad_norm": 1.4435440301895142, | |
| "learning_rate": 2.529305230451666e-05, | |
| "loss": 1.8795, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.505194067955017, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.9938, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.5186567164179103, | |
| "grad_norm": 1.3251738548278809, | |
| "learning_rate": 2.4706947695483348e-05, | |
| "loss": 1.956, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.5373134328358207, | |
| "grad_norm": 1.4197183847427368, | |
| "learning_rate": 2.441393565991849e-05, | |
| "loss": 1.906, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.5559701492537314, | |
| "grad_norm": 1.4905989170074463, | |
| "learning_rate": 2.4121004156723802e-05, | |
| "loss": 1.9073, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.574626865671642, | |
| "grad_norm": 1.3891818523406982, | |
| "learning_rate": 2.3828193438251497e-05, | |
| "loss": 2.1399, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.593283582089552, | |
| "grad_norm": 1.6372982263565063, | |
| "learning_rate": 2.3535543740256536e-05, | |
| "loss": 1.873, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.611940298507463, | |
| "grad_norm": 1.5683703422546387, | |
| "learning_rate": 2.3243095276367685e-05, | |
| "loss": 1.8899, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.6305970149253732, | |
| "grad_norm": 1.585425615310669, | |
| "learning_rate": 2.2950888232561672e-05, | |
| "loss": 2.0511, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.6492537313432836, | |
| "grad_norm": 1.3682692050933838, | |
| "learning_rate": 2.2658962761641232e-05, | |
| "loss": 2.0364, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.667910447761194, | |
| "grad_norm": 1.7755306959152222, | |
| "learning_rate": 2.23673589777175e-05, | |
| "loss": 2.0033, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.6865671641791042, | |
| "grad_norm": 1.4118067026138306, | |
| "learning_rate": 2.207611695069794e-05, | |
| "loss": 2.102, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.705223880597015, | |
| "grad_norm": 1.5786772966384888, | |
| "learning_rate": 2.17852767007802e-05, | |
| "loss": 1.9894, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.7238805970149254, | |
| "grad_norm": 1.4233230352401733, | |
| "learning_rate": 2.1494878192952855e-05, | |
| "loss": 1.9355, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.7425373134328357, | |
| "grad_norm": 1.5830904245376587, | |
| "learning_rate": 2.1204961331503787e-05, | |
| "loss": 1.9399, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.7611940298507465, | |
| "grad_norm": 1.2974706888198853, | |
| "learning_rate": 2.0915565954536744e-05, | |
| "loss": 1.9814, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.779850746268657, | |
| "grad_norm": 1.2366008758544922, | |
| "learning_rate": 2.0626731828497225e-05, | |
| "loss": 1.9275, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.798507462686567, | |
| "grad_norm": 1.5165388584136963, | |
| "learning_rate": 2.0338498642707977e-05, | |
| "loss": 1.9444, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.8171641791044775, | |
| "grad_norm": 1.429136037826538, | |
| "learning_rate": 2.005090600391526e-05, | |
| "loss": 1.9831, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.835820895522388, | |
| "grad_norm": 1.4274283647537231, | |
| "learning_rate": 1.9763993430846395e-05, | |
| "loss": 2.0005, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.8544776119402986, | |
| "grad_norm": 1.502812147140503, | |
| "learning_rate": 1.947780034877938e-05, | |
| "loss": 2.0224, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.873134328358209, | |
| "grad_norm": 1.556489109992981, | |
| "learning_rate": 1.9192366084125425e-05, | |
| "loss": 1.9519, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.8917910447761193, | |
| "grad_norm": 1.467826008796692, | |
| "learning_rate": 1.890772985902496e-05, | |
| "loss": 1.9947, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.91044776119403, | |
| "grad_norm": 1.6837282180786133, | |
| "learning_rate": 1.8623930785958092e-05, | |
| "loss": 1.9335, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.9291044776119404, | |
| "grad_norm": 1.446560025215149, | |
| "learning_rate": 1.8341007862370056e-05, | |
| "loss": 1.9258, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.9477611940298507, | |
| "grad_norm": 1.453008770942688, | |
| "learning_rate": 1.8058999965312484e-05, | |
| "loss": 1.9039, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.966417910447761, | |
| "grad_norm": 1.3427950143814087, | |
| "learning_rate": 1.777794584610124e-05, | |
| "loss": 1.8156, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.9850746268656714, | |
| "grad_norm": 1.7210839986801147, | |
| "learning_rate": 1.749788412499149e-05, | |
| "loss": 2.0007, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.003731343283582, | |
| "grad_norm": 1.8247441053390503, | |
| "learning_rate": 1.721885328587083e-05, | |
| "loss": 1.8995, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.0223880597014925, | |
| "grad_norm": 1.3744760751724243, | |
| "learning_rate": 1.694089167097116e-05, | |
| "loss": 1.9604, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.041044776119403, | |
| "grad_norm": 1.1527031660079956, | |
| "learning_rate": 1.6664037475599923e-05, | |
| "loss": 1.8479, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.0597014925373136, | |
| "grad_norm": 1.412294626235962, | |
| "learning_rate": 1.638832874289168e-05, | |
| "loss": 1.9622, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.078358208955224, | |
| "grad_norm": 1.5206471681594849, | |
| "learning_rate": 1.611380335858047e-05, | |
| "loss": 1.8965, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.0970149253731343, | |
| "grad_norm": 1.426445484161377, | |
| "learning_rate": 1.5840499045793843e-05, | |
| "loss": 1.9118, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.1156716417910446, | |
| "grad_norm": 1.556396245956421, | |
| "learning_rate": 1.5568453359869334e-05, | |
| "loss": 1.8189, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.1343283582089554, | |
| "grad_norm": 1.5185908079147339, | |
| "learning_rate": 1.5297703683193752e-05, | |
| "loss": 1.9363, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.1529850746268657, | |
| "grad_norm": 1.4425839185714722, | |
| "learning_rate": 1.502828722006655e-05, | |
| "loss": 1.9708, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.171641791044776, | |
| "grad_norm": 1.6175637245178223, | |
| "learning_rate": 1.4760240991587337e-05, | |
| "loss": 1.9008, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.1902985074626864, | |
| "grad_norm": 1.5075782537460327, | |
| "learning_rate": 1.4493601830568887e-05, | |
| "loss": 1.9626, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.208955223880597, | |
| "grad_norm": 1.7610998153686523, | |
| "learning_rate": 1.4228406376475742e-05, | |
| "loss": 1.9749, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.2276119402985075, | |
| "grad_norm": 1.538076400756836, | |
| "learning_rate": 1.396469107038956e-05, | |
| "loss": 1.9565, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.246268656716418, | |
| "grad_norm": 1.4104888439178467, | |
| "learning_rate": 1.3702492150001659e-05, | |
| "loss": 1.9042, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.264925373134328, | |
| "grad_norm": 1.5483851432800293, | |
| "learning_rate": 1.34418456446335e-05, | |
| "loss": 1.8595, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.283582089552239, | |
| "grad_norm": 1.8045192956924438, | |
| "learning_rate": 1.3182787370285865e-05, | |
| "loss": 1.8968, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.3022388059701493, | |
| "grad_norm": 1.5665298700332642, | |
| "learning_rate": 1.292535292471726e-05, | |
| "loss": 1.8853, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.3208955223880596, | |
| "grad_norm": 1.4902681112289429, | |
| "learning_rate": 1.2669577682552319e-05, | |
| "loss": 1.8916, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.33955223880597, | |
| "grad_norm": 1.3823623657226562, | |
| "learning_rate": 1.2415496790421011e-05, | |
| "loss": 1.8614, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.3582089552238807, | |
| "grad_norm": 1.4400016069412231, | |
| "learning_rate": 1.2163145162128947e-05, | |
| "loss": 1.9052, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.376865671641791, | |
| "grad_norm": 1.7787601947784424, | |
| "learning_rate": 1.1912557473859895e-05, | |
| "loss": 2.0061, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.3955223880597014, | |
| "grad_norm": 1.5302358865737915, | |
| "learning_rate": 1.1663768159410748e-05, | |
| "loss": 1.9656, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.4141791044776117, | |
| "grad_norm": 1.6571131944656372, | |
| "learning_rate": 1.1416811405459993e-05, | |
| "loss": 1.9289, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.4328358208955225, | |
| "grad_norm": 1.8324801921844482, | |
| "learning_rate": 1.1171721146870015e-05, | |
| "loss": 1.8982, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.451492537313433, | |
| "grad_norm": 1.5971417427062988, | |
| "learning_rate": 1.0928531062024017e-05, | |
| "loss": 1.9105, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.470149253731343, | |
| "grad_norm": 1.5357367992401123, | |
| "learning_rate": 1.0687274568198208e-05, | |
| "loss": 1.9997, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.4888059701492535, | |
| "grad_norm": 1.6085304021835327, | |
| "learning_rate": 1.0447984816969874e-05, | |
| "loss": 1.9139, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.5074626865671643, | |
| "grad_norm": 1.3676837682724, | |
| "learning_rate": 1.021069468966194e-05, | |
| "loss": 1.9389, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.5261194029850746, | |
| "grad_norm": 1.6692901849746704, | |
| "learning_rate": 9.975436792824691e-06, | |
| "loss": 1.835, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.544776119402985, | |
| "grad_norm": 1.579232931137085, | |
| "learning_rate": 9.742243453755202e-06, | |
| "loss": 1.822, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.5634328358208958, | |
| "grad_norm": 1.587336778640747, | |
| "learning_rate": 9.5111467160552e-06, | |
| "loss": 1.9564, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.582089552238806, | |
| "grad_norm": 1.6467562913894653, | |
| "learning_rate": 9.282178335227884e-06, | |
| "loss": 1.9029, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.6007462686567164, | |
| "grad_norm": 1.3579665422439575, | |
| "learning_rate": 9.05536977431431e-06, | |
| "loss": 2.0225, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.6194029850746268, | |
| "grad_norm": 1.641695261001587, | |
| "learning_rate": 8.830752199570033e-06, | |
| "loss": 1.939, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.638059701492537, | |
| "grad_norm": 1.5209190845489502, | |
| "learning_rate": 8.608356476182424e-06, | |
| "loss": 2.01, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.656716417910448, | |
| "grad_norm": 1.769853115081787, | |
| "learning_rate": 8.38821316402946e-06, | |
| "loss": 1.9972, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.675373134328358, | |
| "grad_norm": 1.7627779245376587, | |
| "learning_rate": 8.170352513480408e-06, | |
| "loss": 1.8508, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.6940298507462686, | |
| "grad_norm": 1.8104168176651, | |
| "learning_rate": 7.954804461239053e-06, | |
| "loss": 1.994, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.7126865671641793, | |
| "grad_norm": 1.6434147357940674, | |
| "learning_rate": 7.741598626230079e-06, | |
| "loss": 1.9354, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 3.7313432835820897, | |
| "grad_norm": 1.871103286743164, | |
| "learning_rate": 7.530764305528959e-06, | |
| "loss": 2.004, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 1.7736434936523438, | |
| "learning_rate": 7.3223304703363135e-06, | |
| "loss": 2.0723, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 3.7686567164179103, | |
| "grad_norm": 1.3735415935516357, | |
| "learning_rate": 7.116325761996817e-06, | |
| "loss": 1.9994, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.7873134328358207, | |
| "grad_norm": 1.5601403713226318, | |
| "learning_rate": 6.91277848806356e-06, | |
| "loss": 2.0251, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 3.8059701492537314, | |
| "grad_norm": 1.7761123180389404, | |
| "learning_rate": 6.711716618408281e-06, | |
| "loss": 1.9377, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.824626865671642, | |
| "grad_norm": 1.8018087148666382, | |
| "learning_rate": 6.513167781377885e-06, | |
| "loss": 1.9374, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.843283582089552, | |
| "grad_norm": 1.8684526681900024, | |
| "learning_rate": 6.317159259998073e-06, | |
| "loss": 1.8859, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.861940298507463, | |
| "grad_norm": 1.4688338041305542, | |
| "learning_rate": 6.123717988224237e-06, | |
| "loss": 1.8078, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 3.8805970149253732, | |
| "grad_norm": 1.566017508506775, | |
| "learning_rate": 5.932870547240454e-06, | |
| "loss": 1.82, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.8992537313432836, | |
| "grad_norm": 1.6024701595306396, | |
| "learning_rate": 5.74464316180689e-06, | |
| "loss": 1.9388, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 3.917910447761194, | |
| "grad_norm": 1.615678310394287, | |
| "learning_rate": 5.559061696656198e-06, | |
| "loss": 1.8754, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.9365671641791042, | |
| "grad_norm": 1.7218761444091797, | |
| "learning_rate": 5.37615165293942e-06, | |
| "loss": 1.8457, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 3.955223880597015, | |
| "grad_norm": 1.5426040887832642, | |
| "learning_rate": 5.1959381647217666e-06, | |
| "loss": 1.944, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.9738805970149254, | |
| "grad_norm": 1.607737421989441, | |
| "learning_rate": 5.018445995528931e-06, | |
| "loss": 1.8135, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 3.9925373134328357, | |
| "grad_norm": 1.6999655961990356, | |
| "learning_rate": 4.843699534944257e-06, | |
| "loss": 1.8664, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.0111940298507465, | |
| "grad_norm": 1.7595158815383911, | |
| "learning_rate": 4.671722795257327e-06, | |
| "loss": 1.8825, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.029850746268656, | |
| "grad_norm": 1.6721343994140625, | |
| "learning_rate": 4.502539408164386e-06, | |
| "loss": 1.7858, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.048507462686567, | |
| "grad_norm": 1.8451169729232788, | |
| "learning_rate": 4.336172621521034e-06, | |
| "loss": 1.9469, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.067164179104478, | |
| "grad_norm": 1.5332179069519043, | |
| "learning_rate": 4.1726452961477146e-06, | |
| "loss": 1.8153, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.085820895522388, | |
| "grad_norm": 1.5114800930023193, | |
| "learning_rate": 4.01197990268834e-06, | |
| "loss": 1.92, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.104477611940299, | |
| "grad_norm": 1.642686367034912, | |
| "learning_rate": 3.8541985185225645e-06, | |
| "loss": 1.9352, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.123134328358209, | |
| "grad_norm": 2.106177568435669, | |
| "learning_rate": 3.6993228247320877e-06, | |
| "loss": 1.9699, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.141791044776119, | |
| "grad_norm": 1.5525238513946533, | |
| "learning_rate": 3.547374103121398e-06, | |
| "loss": 1.7852, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.16044776119403, | |
| "grad_norm": 1.49571692943573, | |
| "learning_rate": 3.398373233293378e-06, | |
| "loss": 1.8704, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.17910447761194, | |
| "grad_norm": 1.4679092168807983, | |
| "learning_rate": 3.252340689780245e-06, | |
| "loss": 1.8484, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.197761194029851, | |
| "grad_norm": 1.7375657558441162, | |
| "learning_rate": 3.1092965392300417e-06, | |
| "loss": 1.884, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.2164179104477615, | |
| "grad_norm": 1.5738627910614014, | |
| "learning_rate": 2.969260437649293e-06, | |
| "loss": 1.9437, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.235074626865671, | |
| "grad_norm": 1.9191651344299316, | |
| "learning_rate": 2.8322516277019624e-06, | |
| "loss": 1.9058, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.253731343283582, | |
| "grad_norm": 1.5124653577804565, | |
| "learning_rate": 2.6982889360653377e-06, | |
| "loss": 1.844, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.272388059701493, | |
| "grad_norm": 1.4586025476455688, | |
| "learning_rate": 2.5673907708429976e-06, | |
| "loss": 1.8923, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.291044776119403, | |
| "grad_norm": 1.423386812210083, | |
| "learning_rate": 2.4395751190352924e-06, | |
| "loss": 1.8745, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.309701492537314, | |
| "grad_norm": 1.5810245275497437, | |
| "learning_rate": 2.3148595440677405e-06, | |
| "loss": 1.9122, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.3283582089552235, | |
| "grad_norm": 1.562049388885498, | |
| "learning_rate": 2.1932611833775846e-06, | |
| "loss": 1.9031, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.347014925373134, | |
| "grad_norm": 1.7122998237609863, | |
| "learning_rate": 2.074796746058896e-06, | |
| "loss": 1.9303, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.365671641791045, | |
| "grad_norm": 1.9016716480255127, | |
| "learning_rate": 1.9594825105665654e-06, | |
| "loss": 1.9301, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.384328358208955, | |
| "grad_norm": 1.3229435682296753, | |
| "learning_rate": 1.847334322479413e-06, | |
| "loss": 2.0382, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.402985074626866, | |
| "grad_norm": 1.5966112613677979, | |
| "learning_rate": 1.738367592322837e-06, | |
| "loss": 1.8987, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.4216417910447765, | |
| "grad_norm": 1.6126375198364258, | |
| "learning_rate": 1.6325972934512018e-06, | |
| "loss": 1.7628, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.440298507462686, | |
| "grad_norm": 1.605992317199707, | |
| "learning_rate": 1.5300379599903409e-06, | |
| "loss": 1.9645, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.458955223880597, | |
| "grad_norm": 1.3877547979354858, | |
| "learning_rate": 1.4307036848403648e-06, | |
| "loss": 1.8797, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.477611940298507, | |
| "grad_norm": 1.6082713603973389, | |
| "learning_rate": 1.3346081177391472e-06, | |
| "loss": 1.8971, | |
| "step": 1200 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1340, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2490931612756214e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |