| { | |
| "best_metric": 1.454202651977539, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-500", | |
| "epoch": 0.06560818790185015, | |
| "eval_steps": 100, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0001312163758037003, | |
| "grad_norm": 4.502290725708008, | |
| "learning_rate": 5e-06, | |
| "loss": 2.7959, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0001312163758037003, | |
| "eval_loss": 3.9797542095184326, | |
| "eval_runtime": 1367.0112, | |
| "eval_samples_per_second": 9.39, | |
| "eval_steps_per_second": 2.347, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0002624327516074006, | |
| "grad_norm": 4.730992794036865, | |
| "learning_rate": 1e-05, | |
| "loss": 3.0455, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0003936491274111009, | |
| "grad_norm": 4.7559990882873535, | |
| "learning_rate": 1.5e-05, | |
| "loss": 3.0463, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0005248655032148012, | |
| "grad_norm": 4.74609899520874, | |
| "learning_rate": 2e-05, | |
| "loss": 3.0803, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0006560818790185015, | |
| "grad_norm": 4.335945129394531, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.885, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0007872982548222018, | |
| "grad_norm": 4.600176811218262, | |
| "learning_rate": 3e-05, | |
| "loss": 3.0, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0009185146306259021, | |
| "grad_norm": 4.308350563049316, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.9882, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0010497310064296024, | |
| "grad_norm": 4.452569484710693, | |
| "learning_rate": 4e-05, | |
| "loss": 2.8844, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0011809473822333026, | |
| "grad_norm": 5.113404750823975, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.8185, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.001312163758037003, | |
| "grad_norm": 4.601175785064697, | |
| "learning_rate": 5e-05, | |
| "loss": 2.7809, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0014433801338407032, | |
| "grad_norm": 4.8205156326293945, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.514, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0015745965096444037, | |
| "grad_norm": 4.931856155395508, | |
| "learning_rate": 6e-05, | |
| "loss": 2.6665, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0017058128854481039, | |
| "grad_norm": 6.019423007965088, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.6067, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0018370292612518043, | |
| "grad_norm": 4.860208034515381, | |
| "learning_rate": 7e-05, | |
| "loss": 2.7381, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0019682456370555047, | |
| "grad_norm": 5.671669960021973, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.4756, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.002099462012859205, | |
| "grad_norm": 4.821615219116211, | |
| "learning_rate": 8e-05, | |
| "loss": 2.5178, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.002230678388662905, | |
| "grad_norm": 5.072446823120117, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.4209, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0023618947644666053, | |
| "grad_norm": 5.528189659118652, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4476, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.002493111140270306, | |
| "grad_norm": 4.594228267669678, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.2791, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.002624327516074006, | |
| "grad_norm": 5.120175838470459, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2744, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0027555438918777063, | |
| "grad_norm": 4.523880958557129, | |
| "learning_rate": 9.999892908320647e-05, | |
| "loss": 2.0823, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0028867602676814065, | |
| "grad_norm": 5.103494167327881, | |
| "learning_rate": 9.999571637870036e-05, | |
| "loss": 2.2586, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.003017976643485107, | |
| "grad_norm": 4.991058826446533, | |
| "learning_rate": 9.999036202410325e-05, | |
| "loss": 2.1277, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0031491930192888073, | |
| "grad_norm": 4.875056743621826, | |
| "learning_rate": 9.998286624877786e-05, | |
| "loss": 2.1464, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0032804093950925075, | |
| "grad_norm": 4.341136932373047, | |
| "learning_rate": 9.997322937381829e-05, | |
| "loss": 2.1301, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0034116257708962077, | |
| "grad_norm": 4.211912155151367, | |
| "learning_rate": 9.996145181203615e-05, | |
| "loss": 2.0398, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0035428421466999083, | |
| "grad_norm": 4.3173651695251465, | |
| "learning_rate": 9.994753406794301e-05, | |
| "loss": 2.063, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0036740585225036085, | |
| "grad_norm": 4.31871223449707, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 1.9695, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0038052748983073087, | |
| "grad_norm": 5.267412185668945, | |
| "learning_rate": 9.991328050923581e-05, | |
| "loss": 2.017, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.003936491274111009, | |
| "grad_norm": 4.116550922393799, | |
| "learning_rate": 9.989294616193017e-05, | |
| "loss": 1.674, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.004067707649914709, | |
| "grad_norm": 4.253255367279053, | |
| "learning_rate": 9.98704745668676e-05, | |
| "loss": 1.8298, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.00419892402571841, | |
| "grad_norm": 7.945616245269775, | |
| "learning_rate": 9.98458666866564e-05, | |
| "loss": 1.8191, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.00433014040152211, | |
| "grad_norm": 4.653039455413818, | |
| "learning_rate": 9.981912357541627e-05, | |
| "loss": 1.786, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.00446135677732581, | |
| "grad_norm": 5.207880020141602, | |
| "learning_rate": 9.97902463787331e-05, | |
| "loss": 1.9124, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.004592573153129511, | |
| "grad_norm": 4.892001152038574, | |
| "learning_rate": 9.975923633360985e-05, | |
| "loss": 1.5737, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0047237895289332105, | |
| "grad_norm": 4.775864124298096, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.7467, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.004855005904736911, | |
| "grad_norm": 5.307229518890381, | |
| "learning_rate": 9.969082310281891e-05, | |
| "loss": 1.5144, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.004986222280540612, | |
| "grad_norm": 5.3355560302734375, | |
| "learning_rate": 9.965342284774632e-05, | |
| "loss": 1.576, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0051174386563443116, | |
| "grad_norm": 4.783757209777832, | |
| "learning_rate": 9.961389560529836e-05, | |
| "loss": 1.3935, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.005248655032148012, | |
| "grad_norm": 4.547131061553955, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 1.4015, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.005379871407951712, | |
| "grad_norm": 6.1063055992126465, | |
| "learning_rate": 9.952846702217886e-05, | |
| "loss": 1.402, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.005511087783755413, | |
| "grad_norm": 4.746460437774658, | |
| "learning_rate": 9.948256934098352e-05, | |
| "loss": 1.1739, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.005642304159559113, | |
| "grad_norm": 4.250523567199707, | |
| "learning_rate": 9.943455199120837e-05, | |
| "loss": 1.2131, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.005773520535362813, | |
| "grad_norm": 5.288604736328125, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 1.1556, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.005904736911166514, | |
| "grad_norm": 5.479332447052002, | |
| "learning_rate": 9.933216660424395e-05, | |
| "loss": 1.2704, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.006035953286970214, | |
| "grad_norm": 4.803893566131592, | |
| "learning_rate": 9.927780295290389e-05, | |
| "loss": 1.0887, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.006167169662773914, | |
| "grad_norm": 4.377706050872803, | |
| "learning_rate": 9.922132840449459e-05, | |
| "loss": 0.8161, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.006298386038577615, | |
| "grad_norm": 4.871409893035889, | |
| "learning_rate": 9.916274537819775e-05, | |
| "loss": 0.8478, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.006429602414381314, | |
| "grad_norm": 4.593444347381592, | |
| "learning_rate": 9.91020563835152e-05, | |
| "loss": 0.9841, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.006560818790185015, | |
| "grad_norm": 6.342800617218018, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.8137, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.006692035165988716, | |
| "grad_norm": 3.515503406524658, | |
| "learning_rate": 9.897437097795257e-05, | |
| "loss": 2.5555, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.006823251541792415, | |
| "grad_norm": 2.8081209659576416, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 2.4244, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.006954467917596116, | |
| "grad_norm": 2.0723488330841064, | |
| "learning_rate": 9.883829406604363e-05, | |
| "loss": 2.255, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.007085684293399817, | |
| "grad_norm": 1.63498055934906, | |
| "learning_rate": 9.876711602542563e-05, | |
| "loss": 2.219, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0072169006692035164, | |
| "grad_norm": 1.5121456384658813, | |
| "learning_rate": 9.869384896386668e-05, | |
| "loss": 2.2263, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.007348117045007217, | |
| "grad_norm": 1.7543376684188843, | |
| "learning_rate": 9.861849601988383e-05, | |
| "loss": 2.2319, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.007479333420810917, | |
| "grad_norm": 2.030773639678955, | |
| "learning_rate": 9.854106042134641e-05, | |
| "loss": 2.299, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.0076105497966146175, | |
| "grad_norm": 2.207674264907837, | |
| "learning_rate": 9.846154548533773e-05, | |
| "loss": 2.3749, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.007741766172418318, | |
| "grad_norm": 2.266341209411621, | |
| "learning_rate": 9.837995461801299e-05, | |
| "loss": 2.3051, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.007872982548222019, | |
| "grad_norm": 2.5948050022125244, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 2.2048, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.008004198924025718, | |
| "grad_norm": 3.1021060943603516, | |
| "learning_rate": 9.821055915851647e-05, | |
| "loss": 2.3619, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.008135415299829418, | |
| "grad_norm": 3.0172550678253174, | |
| "learning_rate": 9.812276182268236e-05, | |
| "loss": 2.198, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.008266631675633119, | |
| "grad_norm": 4.317417144775391, | |
| "learning_rate": 9.803290306789676e-05, | |
| "loss": 2.1099, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.00839784805143682, | |
| "grad_norm": 4.255865097045898, | |
| "learning_rate": 9.794098674340965e-05, | |
| "loss": 2.2599, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.00852906442724052, | |
| "grad_norm": 3.712496280670166, | |
| "learning_rate": 9.784701678661045e-05, | |
| "loss": 2.0834, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.00866028080304422, | |
| "grad_norm": 5.127752304077148, | |
| "learning_rate": 9.775099722285935e-05, | |
| "loss": 2.082, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.00879149717884792, | |
| "grad_norm": 3.70094633102417, | |
| "learning_rate": 9.765293216531486e-05, | |
| "loss": 2.2118, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.00892271355465162, | |
| "grad_norm": 3.54309344291687, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.8701, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.009053929930455321, | |
| "grad_norm": 4.72603702545166, | |
| "learning_rate": 9.74506824594107e-05, | |
| "loss": 2.1741, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.009185146306259022, | |
| "grad_norm": 4.20818567276001, | |
| "learning_rate": 9.73465064747553e-05, | |
| "loss": 1.9426, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.009316362682062722, | |
| "grad_norm": 3.797800064086914, | |
| "learning_rate": 9.724030232334391e-05, | |
| "loss": 1.8893, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.009447579057866421, | |
| "grad_norm": 4.630662441253662, | |
| "learning_rate": 9.713207455460894e-05, | |
| "loss": 1.91, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.009578795433670122, | |
| "grad_norm": 4.435169696807861, | |
| "learning_rate": 9.702182780466775e-05, | |
| "loss": 1.8245, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.009710011809473822, | |
| "grad_norm": 4.108457088470459, | |
| "learning_rate": 9.690956679612421e-05, | |
| "loss": 1.8773, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.009841228185277523, | |
| "grad_norm": 3.7498910427093506, | |
| "learning_rate": 9.67952963378663e-05, | |
| "loss": 1.5919, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.009972444561081224, | |
| "grad_norm": 3.4361634254455566, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 1.5667, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.010103660936884922, | |
| "grad_norm": 3.993279218673706, | |
| "learning_rate": 9.656074673794018e-05, | |
| "loss": 1.8395, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.010234877312688623, | |
| "grad_norm": 3.958552598953247, | |
| "learning_rate": 9.644047764359622e-05, | |
| "loss": 1.7246, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.010366093688492324, | |
| "grad_norm": 4.14636754989624, | |
| "learning_rate": 9.631821919375591e-05, | |
| "loss": 1.6861, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.010497310064296024, | |
| "grad_norm": 4.084466934204102, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 1.4632, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.010628526440099725, | |
| "grad_norm": 3.8091354370117188, | |
| "learning_rate": 9.606775526115963e-05, | |
| "loss": 1.6115, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.010759742815903424, | |
| "grad_norm": 3.6534502506256104, | |
| "learning_rate": 9.593956050744492e-05, | |
| "loss": 1.4976, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.010890959191707125, | |
| "grad_norm": 4.479729652404785, | |
| "learning_rate": 9.580939785585681e-05, | |
| "loss": 1.6682, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.011022175567510825, | |
| "grad_norm": 4.333984375, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.7758, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.011153391943314526, | |
| "grad_norm": 5.193741321563721, | |
| "learning_rate": 9.554319124605879e-05, | |
| "loss": 1.382, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.011284608319118226, | |
| "grad_norm": 4.566789627075195, | |
| "learning_rate": 9.540715869125407e-05, | |
| "loss": 1.5179, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.011415824694921927, | |
| "grad_norm": 4.185393810272217, | |
| "learning_rate": 9.526918104489777e-05, | |
| "loss": 1.3469, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.011547041070725626, | |
| "grad_norm": 4.056511878967285, | |
| "learning_rate": 9.512926421749304e-05, | |
| "loss": 1.2551, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.011678257446529327, | |
| "grad_norm": 4.229364395141602, | |
| "learning_rate": 9.498741420261108e-05, | |
| "loss": 1.4315, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.011809473822333027, | |
| "grad_norm": 3.8965346813201904, | |
| "learning_rate": 9.484363707663442e-05, | |
| "loss": 1.4387, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.011940690198136728, | |
| "grad_norm": 4.694686412811279, | |
| "learning_rate": 9.469793899849661e-05, | |
| "loss": 1.5771, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.012071906573940428, | |
| "grad_norm": 5.034073352813721, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 1.6219, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.012203122949744127, | |
| "grad_norm": 4.071391582489014, | |
| "learning_rate": 9.440080503264037e-05, | |
| "loss": 1.0868, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.012334339325547828, | |
| "grad_norm": 4.7768168449401855, | |
| "learning_rate": 9.42493818731521e-05, | |
| "loss": 1.2184, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.012465555701351529, | |
| "grad_norm": 4.308196067810059, | |
| "learning_rate": 9.409606321741775e-05, | |
| "loss": 0.852, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.01259677207715523, | |
| "grad_norm": 3.9685397148132324, | |
| "learning_rate": 9.394085563309827e-05, | |
| "loss": 1.0385, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.01272798845295893, | |
| "grad_norm": 3.9740753173828125, | |
| "learning_rate": 9.378376576876999e-05, | |
| "loss": 0.9448, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.012859204828762629, | |
| "grad_norm": 3.3262381553649902, | |
| "learning_rate": 9.362480035363986e-05, | |
| "loss": 0.6915, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01299042120456633, | |
| "grad_norm": 5.052366256713867, | |
| "learning_rate": 9.34639661972572e-05, | |
| "loss": 0.8459, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.01312163758037003, | |
| "grad_norm": 4.026389122009277, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.6136, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01312163758037003, | |
| "eval_loss": 1.995261549949646, | |
| "eval_runtime": 1352.3203, | |
| "eval_samples_per_second": 9.492, | |
| "eval_steps_per_second": 2.373, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01325285395617373, | |
| "grad_norm": 3.1780271530151367, | |
| "learning_rate": 9.31367192988896e-05, | |
| "loss": 2.3667, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.013384070331977431, | |
| "grad_norm": 2.605407238006592, | |
| "learning_rate": 9.297032057507264e-05, | |
| "loss": 2.2015, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.013515286707781132, | |
| "grad_norm": 1.9591269493103027, | |
| "learning_rate": 9.280208114573859e-05, | |
| "loss": 2.0853, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.01364650308358483, | |
| "grad_norm": 1.4972039461135864, | |
| "learning_rate": 9.263200821770461e-05, | |
| "loss": 2.0733, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.013777719459388531, | |
| "grad_norm": 1.6665318012237549, | |
| "learning_rate": 9.246010907632895e-05, | |
| "loss": 2.0467, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.013908935835192232, | |
| "grad_norm": 1.6789878606796265, | |
| "learning_rate": 9.228639108519868e-05, | |
| "loss": 2.1175, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.014040152210995933, | |
| "grad_norm": 2.0789976119995117, | |
| "learning_rate": 9.211086168581433e-05, | |
| "loss": 2.2926, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.014171368586799633, | |
| "grad_norm": 2.8981029987335205, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 2.1756, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.014302584962603332, | |
| "grad_norm": 2.743936538696289, | |
| "learning_rate": 9.175439881593716e-05, | |
| "loss": 2.1466, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.014433801338407033, | |
| "grad_norm": 3.987334728240967, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 2.2559, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.014565017714210734, | |
| "grad_norm": 3.1965720653533936, | |
| "learning_rate": 9.139078154477512e-05, | |
| "loss": 2.2562, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.014696234090014434, | |
| "grad_norm": 3.892089605331421, | |
| "learning_rate": 9.120630943110077e-05, | |
| "loss": 2.1416, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.014827450465818135, | |
| "grad_norm": 3.2245702743530273, | |
| "learning_rate": 9.102007217627568e-05, | |
| "loss": 2.0573, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.014958666841621834, | |
| "grad_norm": 3.0087993144989014, | |
| "learning_rate": 9.083207775808396e-05, | |
| "loss": 1.9645, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.015089883217425534, | |
| "grad_norm": 2.868622303009033, | |
| "learning_rate": 9.064233422958077e-05, | |
| "loss": 2.0366, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.015221099593229235, | |
| "grad_norm": 3.084916830062866, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.0537, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.015352315969032936, | |
| "grad_norm": 3.2997586727142334, | |
| "learning_rate": 9.025763242814291e-05, | |
| "loss": 2.0362, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.015483532344836636, | |
| "grad_norm": 4.677249431610107, | |
| "learning_rate": 9.006269063455304e-05, | |
| "loss": 1.8742, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.015614748720640335, | |
| "grad_norm": 4.065469741821289, | |
| "learning_rate": 8.986603268863536e-05, | |
| "loss": 1.7975, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.015745965096444037, | |
| "grad_norm": 3.6069369316101074, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 1.7775, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.015877181472247738, | |
| "grad_norm": 4.557692527770996, | |
| "learning_rate": 8.94676021096575e-05, | |
| "loss": 1.8745, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.016008397848051435, | |
| "grad_norm": 3.5455398559570312, | |
| "learning_rate": 8.926584654403724e-05, | |
| "loss": 1.7982, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.016139614223855136, | |
| "grad_norm": 3.4866509437561035, | |
| "learning_rate": 8.906240896023794e-05, | |
| "loss": 1.7942, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.016270830599658836, | |
| "grad_norm": 3.9673101902008057, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 1.6027, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.016402046975462537, | |
| "grad_norm": 4.192202568054199, | |
| "learning_rate": 8.865052266813685e-05, | |
| "loss": 1.7069, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.016533263351266238, | |
| "grad_norm": 3.4886741638183594, | |
| "learning_rate": 8.844209160367299e-05, | |
| "loss": 1.6127, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.01666447972706994, | |
| "grad_norm": 4.3129448890686035, | |
| "learning_rate": 8.823201380795001e-05, | |
| "loss": 1.8012, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.01679569610287364, | |
| "grad_norm": 4.046361923217773, | |
| "learning_rate": 8.802029828000156e-05, | |
| "loss": 1.6391, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.01692691247867734, | |
| "grad_norm": 3.780560255050659, | |
| "learning_rate": 8.780695408901613e-05, | |
| "loss": 1.8057, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.01705812885448104, | |
| "grad_norm": 4.130734443664551, | |
| "learning_rate": 8.759199037394887e-05, | |
| "loss": 1.7628, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.01718934523028474, | |
| "grad_norm": 4.335687160491943, | |
| "learning_rate": 8.737541634312985e-05, | |
| "loss": 1.7281, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.01732056160608844, | |
| "grad_norm": 3.8232834339141846, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 1.6856, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.01745177798189214, | |
| "grad_norm": 4.297300338745117, | |
| "learning_rate": 8.693747451206232e-05, | |
| "loss": 1.4618, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.01758299435769584, | |
| "grad_norm": 3.4930808544158936, | |
| "learning_rate": 8.671612547178428e-05, | |
| "loss": 1.3184, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.01771421073349954, | |
| "grad_norm": 3.8762755393981934, | |
| "learning_rate": 8.649320363489179e-05, | |
| "loss": 1.4333, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.01784542710930324, | |
| "grad_norm": 4.0097126960754395, | |
| "learning_rate": 8.626871855061438e-05, | |
| "loss": 1.4535, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.01797664348510694, | |
| "grad_norm": 4.139339923858643, | |
| "learning_rate": 8.604267983514594e-05, | |
| "loss": 1.4856, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.018107859860910642, | |
| "grad_norm": 3.6388514041900635, | |
| "learning_rate": 8.581509717123273e-05, | |
| "loss": 1.3928, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.018239076236714342, | |
| "grad_norm": 3.7924964427948, | |
| "learning_rate": 8.558598030775857e-05, | |
| "loss": 1.2458, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.018370292612518043, | |
| "grad_norm": 4.4245758056640625, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 1.4169, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.018501508988321744, | |
| "grad_norm": 3.6946239471435547, | |
| "learning_rate": 8.51231833058426e-05, | |
| "loss": 1.1482, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.018632725364125444, | |
| "grad_norm": 3.711535692214966, | |
| "learning_rate": 8.488952299208401e-05, | |
| "loss": 1.065, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.01876394173992914, | |
| "grad_norm": 3.8962907791137695, | |
| "learning_rate": 8.46543681272818e-05, | |
| "loss": 1.2842, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.018895158115732842, | |
| "grad_norm": 3.2645106315612793, | |
| "learning_rate": 8.44177287846877e-05, | |
| "loss": 0.8512, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.019026374491536543, | |
| "grad_norm": 3.1213479042053223, | |
| "learning_rate": 8.417961510114356e-05, | |
| "loss": 0.7018, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.019157590867340243, | |
| "grad_norm": 4.530689716339111, | |
| "learning_rate": 8.39400372766471e-05, | |
| "loss": 0.9977, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.019288807243143944, | |
| "grad_norm": 3.922245740890503, | |
| "learning_rate": 8.36990055739149e-05, | |
| "loss": 0.6566, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.019420023618947645, | |
| "grad_norm": 4.297173500061035, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.7287, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.019551239994751345, | |
| "grad_norm": 4.700723648071289, | |
| "learning_rate": 8.321262189556409e-05, | |
| "loss": 0.724, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.019682456370555046, | |
| "grad_norm": 4.133970737457275, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 0.5965, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.019813672746358747, | |
| "grad_norm": 2.415865898132324, | |
| "learning_rate": 8.272054740543052e-05, | |
| "loss": 2.1392, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.019944889122162447, | |
| "grad_norm": 2.01712965965271, | |
| "learning_rate": 8.247240241650918e-05, | |
| "loss": 2.1176, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.020076105497966148, | |
| "grad_norm": 1.6205581426620483, | |
| "learning_rate": 8.222286641794488e-05, | |
| "loss": 2.1266, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.020207321873769845, | |
| "grad_norm": 1.8821977376937866, | |
| "learning_rate": 8.197195009902924e-05, | |
| "loss": 2.126, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.020338538249573546, | |
| "grad_norm": 1.4580323696136475, | |
| "learning_rate": 8.171966420818228e-05, | |
| "loss": 2.081, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.020469754625377246, | |
| "grad_norm": 1.5964078903198242, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 2.1827, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.020600971001180947, | |
| "grad_norm": 2.922008752822876, | |
| "learning_rate": 8.121102699725089e-05, | |
| "loss": 2.1915, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.020732187376984648, | |
| "grad_norm": 2.2149009704589844, | |
| "learning_rate": 8.095469746549172e-05, | |
| "loss": 2.1432, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.020863403752788348, | |
| "grad_norm": 2.5889828205108643, | |
| "learning_rate": 8.069704193751832e-05, | |
| "loss": 2.101, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.02099462012859205, | |
| "grad_norm": 3.5003163814544678, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 2.1131, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.02112583650439575, | |
| "grad_norm": 3.9812967777252197, | |
| "learning_rate": 8.017779709767858e-05, | |
| "loss": 2.1469, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.02125705288019945, | |
| "grad_norm": 3.2259273529052734, | |
| "learning_rate": 7.991623002853296e-05, | |
| "loss": 2.0289, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.02138826925600315, | |
| "grad_norm": 4.384939670562744, | |
| "learning_rate": 7.965338144766186e-05, | |
| "loss": 2.2525, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.021519485631806848, | |
| "grad_norm": 4.085958957672119, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.0116, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.02165070200761055, | |
| "grad_norm": 2.951061964035034, | |
| "learning_rate": 7.912388484339012e-05, | |
| "loss": 2.0182, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.02178191838341425, | |
| "grad_norm": 5.261083126068115, | |
| "learning_rate": 7.88572595018617e-05, | |
| "loss": 1.8649, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.02191313475921795, | |
| "grad_norm": 5.297954082489014, | |
| "learning_rate": 7.858939801138061e-05, | |
| "loss": 1.9873, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.02204435113502165, | |
| "grad_norm": 3.1123671531677246, | |
| "learning_rate": 7.832031184624164e-05, | |
| "loss": 1.9712, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.02217556751082535, | |
| "grad_norm": 3.0200247764587402, | |
| "learning_rate": 7.80500125332005e-05, | |
| "loss": 1.7175, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.02230678388662905, | |
| "grad_norm": 4.045231819152832, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 1.6917, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.022438000262432752, | |
| "grad_norm": 3.3924217224121094, | |
| "learning_rate": 7.750582082977467e-05, | |
| "loss": 1.915, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.022569216638236453, | |
| "grad_norm": 3.338836669921875, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 1.7668, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.022700433014040153, | |
| "grad_norm": 3.4897708892822266, | |
| "learning_rate": 7.695691614555003e-05, | |
| "loss": 1.7118, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.022831649389843854, | |
| "grad_norm": 3.5899102687835693, | |
| "learning_rate": 7.668072579578058e-05, | |
| "loss": 1.6634, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.02296286576564755, | |
| "grad_norm": 3.5203874111175537, | |
| "learning_rate": 7.64033925325184e-05, | |
| "loss": 1.8026, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.023094082141451252, | |
| "grad_norm": 3.14801287651062, | |
| "learning_rate": 7.612492823579745e-05, | |
| "loss": 1.6852, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.023225298517254953, | |
| "grad_norm": 3.505002021789551, | |
| "learning_rate": 7.584534483410137e-05, | |
| "loss": 1.6794, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.023356514893058653, | |
| "grad_norm": 3.6999619007110596, | |
| "learning_rate": 7.55646543038526e-05, | |
| "loss": 1.6416, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.023487731268862354, | |
| "grad_norm": 3.438927412033081, | |
| "learning_rate": 7.528286866889924e-05, | |
| "loss": 1.6118, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.023618947644666054, | |
| "grad_norm": 3.4257144927978516, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.4316, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.023750164020469755, | |
| "grad_norm": 3.3115530014038086, | |
| "learning_rate": 7.471606041430723e-05, | |
| "loss": 1.4048, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.023881380396273456, | |
| "grad_norm": 3.7471344470977783, | |
| "learning_rate": 7.443106207484776e-05, | |
| "loss": 1.4955, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.024012596772077156, | |
| "grad_norm": 3.900952100753784, | |
| "learning_rate": 7.414501719000187e-05, | |
| "loss": 1.5621, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.024143813147880857, | |
| "grad_norm": 3.704540491104126, | |
| "learning_rate": 7.385793801298042e-05, | |
| "loss": 1.4024, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.024275029523684554, | |
| "grad_norm": 3.8118085861206055, | |
| "learning_rate": 7.35698368412999e-05, | |
| "loss": 1.4171, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.024406245899488255, | |
| "grad_norm": 4.9681196212768555, | |
| "learning_rate": 7.328072601625557e-05, | |
| "loss": 1.4693, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.024537462275291955, | |
| "grad_norm": 3.6511571407318115, | |
| "learning_rate": 7.2990617922393e-05, | |
| "loss": 1.2867, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.024668678651095656, | |
| "grad_norm": 3.6810758113861084, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 1.1199, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.024799895026899357, | |
| "grad_norm": 3.7081968784332275, | |
| "learning_rate": 7.240745967946113e-05, | |
| "loss": 1.0948, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.024931111402703057, | |
| "grad_norm": 3.9994025230407715, | |
| "learning_rate": 7.211443451095007e-05, | |
| "loss": 1.3183, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.025062327778506758, | |
| "grad_norm": 3.6044254302978516, | |
| "learning_rate": 7.18204620336671e-05, | |
| "loss": 1.2563, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.02519354415431046, | |
| "grad_norm": 3.95975923538208, | |
| "learning_rate": 7.152555484041476e-05, | |
| "loss": 0.9602, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.02532476053011416, | |
| "grad_norm": 3.7815332412719727, | |
| "learning_rate": 7.122972556403567e-05, | |
| "loss": 1.0867, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.02545597690591786, | |
| "grad_norm": 3.6170129776000977, | |
| "learning_rate": 7.09329868768714e-05, | |
| "loss": 0.8663, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.02558719328172156, | |
| "grad_norm": 3.3122055530548096, | |
| "learning_rate": 7.063535149021973e-05, | |
| "loss": 0.8313, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.025718409657525258, | |
| "grad_norm": 4.299932479858398, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.1031, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.025849626033328958, | |
| "grad_norm": 3.6584527492523193, | |
| "learning_rate": 7.003744165515705e-05, | |
| "loss": 0.6661, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.02598084240913266, | |
| "grad_norm": 3.595172166824341, | |
| "learning_rate": 6.973719281921335e-05, | |
| "loss": 0.6774, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.02611205878493636, | |
| "grad_norm": 3.2948033809661865, | |
| "learning_rate": 6.943609850761979e-05, | |
| "loss": 0.6726, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.02624327516074006, | |
| "grad_norm": 4.285879135131836, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.648, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02624327516074006, | |
| "eval_loss": 1.8295915126800537, | |
| "eval_runtime": 1356.6996, | |
| "eval_samples_per_second": 9.461, | |
| "eval_steps_per_second": 2.365, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02637449153654376, | |
| "grad_norm": 2.5337905883789062, | |
| "learning_rate": 6.883142508466054e-05, | |
| "loss": 1.9931, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.02650570791234746, | |
| "grad_norm": 1.4764660596847534, | |
| "learning_rate": 6.852787187549182e-05, | |
| "loss": 2.0433, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.026636924288151162, | |
| "grad_norm": 2.1547956466674805, | |
| "learning_rate": 6.82235249939575e-05, | |
| "loss": 1.9185, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.026768140663954863, | |
| "grad_norm": 2.4460480213165283, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 1.9992, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.026899357039758563, | |
| "grad_norm": 1.4299720525741577, | |
| "learning_rate": 6.761250239606169e-05, | |
| "loss": 2.0654, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.027030573415562264, | |
| "grad_norm": 1.9943609237670898, | |
| "learning_rate": 6.730585285387465e-05, | |
| "loss": 2.0979, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.02716178979136596, | |
| "grad_norm": 2.2427730560302734, | |
| "learning_rate": 6.699846198654971e-05, | |
| "loss": 2.2645, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.02729300616716966, | |
| "grad_norm": 2.437592029571533, | |
| "learning_rate": 6.669034296168855e-05, | |
| "loss": 2.2168, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.027424222542973362, | |
| "grad_norm": 2.5972635746002197, | |
| "learning_rate": 6.638150897808468e-05, | |
| "loss": 2.2821, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.027555438918777063, | |
| "grad_norm": 2.6581530570983887, | |
| "learning_rate": 6.607197326515808e-05, | |
| "loss": 2.1277, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.027686655294580764, | |
| "grad_norm": 3.457130193710327, | |
| "learning_rate": 6.57617490823885e-05, | |
| "loss": 2.0846, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.027817871670384464, | |
| "grad_norm": 2.862334728240967, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.0455, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.027949088046188165, | |
| "grad_norm": 2.7949256896972656, | |
| "learning_rate": 6.513928849212873e-05, | |
| "loss": 1.8195, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.028080304421991865, | |
| "grad_norm": 3.603222608566284, | |
| "learning_rate": 6.482707874877854e-05, | |
| "loss": 2.1141, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.028211520797795566, | |
| "grad_norm": 3.7938568592071533, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 2.0341, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.028342737173599267, | |
| "grad_norm": 3.7415919303894043, | |
| "learning_rate": 6.420076723519614e-05, | |
| "loss": 1.9174, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.028473953549402964, | |
| "grad_norm": 3.2029411792755127, | |
| "learning_rate": 6.388669229406462e-05, | |
| "loss": 1.9153, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.028605169925206664, | |
| "grad_norm": 4.757354259490967, | |
| "learning_rate": 6.357202249325371e-05, | |
| "loss": 2.0522, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.028736386301010365, | |
| "grad_norm": 4.18367862701416, | |
| "learning_rate": 6.32567713121704e-05, | |
| "loss": 1.7395, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.028867602676814066, | |
| "grad_norm": 3.411064863204956, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 1.8052, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.028998819052617766, | |
| "grad_norm": 3.278550624847412, | |
| "learning_rate": 6.26245788507579e-05, | |
| "loss": 1.9089, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.029130035428421467, | |
| "grad_norm": 3.312464714050293, | |
| "learning_rate": 6.230766465144967e-05, | |
| "loss": 1.7548, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.029261251804225168, | |
| "grad_norm": 3.0991432666778564, | |
| "learning_rate": 6.199022323275083e-05, | |
| "loss": 1.6692, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.02939246818002887, | |
| "grad_norm": 3.4286837577819824, | |
| "learning_rate": 6.167226819279528e-05, | |
| "loss": 1.79, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.02952368455583257, | |
| "grad_norm": 3.0509002208709717, | |
| "learning_rate": 6.135381315171867e-05, | |
| "loss": 1.5622, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.02965490093163627, | |
| "grad_norm": 3.1480135917663574, | |
| "learning_rate": 6.103487175107507e-05, | |
| "loss": 1.622, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.02978611730743997, | |
| "grad_norm": 3.6017837524414062, | |
| "learning_rate": 6.071545765325254e-05, | |
| "loss": 1.6357, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.029917333683243667, | |
| "grad_norm": 3.5763702392578125, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.647, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.030048550059047368, | |
| "grad_norm": 3.3560070991516113, | |
| "learning_rate": 6.007526611628086e-05, | |
| "loss": 1.4219, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.03017976643485107, | |
| "grad_norm": 3.4218170642852783, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 1.5417, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.03031098281065477, | |
| "grad_norm": 3.3482179641723633, | |
| "learning_rate": 5.9433348234327765e-05, | |
| "loss": 1.4549, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.03044219918645847, | |
| "grad_norm": 3.298544406890869, | |
| "learning_rate": 5.911177627460739e-05, | |
| "loss": 1.4727, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.03057341556226217, | |
| "grad_norm": 3.360682487487793, | |
| "learning_rate": 5.8789813996717736e-05, | |
| "loss": 1.4171, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.03070463193806587, | |
| "grad_norm": 3.3437387943267822, | |
| "learning_rate": 5.8467475192451226e-05, | |
| "loss": 1.2873, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.03083584831386957, | |
| "grad_norm": 3.6862969398498535, | |
| "learning_rate": 5.814477366972945e-05, | |
| "loss": 1.5642, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.030967064689673272, | |
| "grad_norm": 3.4449985027313232, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 1.2568, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.031098281065476973, | |
| "grad_norm": 4.007747173309326, | |
| "learning_rate": 5.749833777770225e-05, | |
| "loss": 1.1918, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.03122949744128067, | |
| "grad_norm": 3.507702589035034, | |
| "learning_rate": 5.717463109955896e-05, | |
| "loss": 1.2402, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.03136071381708437, | |
| "grad_norm": 3.6946218013763428, | |
| "learning_rate": 5.685061708409841e-05, | |
| "loss": 1.4288, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.031491930192888075, | |
| "grad_norm": 4.057668209075928, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 1.1473, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.03162314656869177, | |
| "grad_norm": 3.8214073181152344, | |
| "learning_rate": 5.6201722572524275e-05, | |
| "loss": 1.1556, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.031754362944495476, | |
| "grad_norm": 3.187711000442505, | |
| "learning_rate": 5.587686987289189e-05, | |
| "loss": 0.8827, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.03188557932029917, | |
| "grad_norm": 7.194174289703369, | |
| "learning_rate": 5.5551765427713884e-05, | |
| "loss": 1.1186, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.03201679569610287, | |
| "grad_norm": 3.3873300552368164, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.9614, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.032148012071906575, | |
| "grad_norm": 3.6875414848327637, | |
| "learning_rate": 5.490085701647805e-05, | |
| "loss": 0.9514, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.03227922844771027, | |
| "grad_norm": 2.922820806503296, | |
| "learning_rate": 5.457508093317013e-05, | |
| "loss": 0.6782, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.032410444823513976, | |
| "grad_norm": 2.960690975189209, | |
| "learning_rate": 5.4249108868622086e-05, | |
| "loss": 0.5822, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.03254166119931767, | |
| "grad_norm": 2.816145658493042, | |
| "learning_rate": 5.392295478639225e-05, | |
| "loss": 0.5743, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.03267287757512138, | |
| "grad_norm": 3.569626569747925, | |
| "learning_rate": 5.359663265783598e-05, | |
| "loss": 0.6907, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.032804093950925074, | |
| "grad_norm": 4.098865032196045, | |
| "learning_rate": 5.327015646150716e-05, | |
| "loss": 0.7811, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.03293531032672878, | |
| "grad_norm": 1.4596612453460693, | |
| "learning_rate": 5.294354018255945e-05, | |
| "loss": 1.9946, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.033066526702532476, | |
| "grad_norm": 1.3144938945770264, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 1.9186, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.03319774307833618, | |
| "grad_norm": 1.3099392652511597, | |
| "learning_rate": 5.228994334682604e-05, | |
| "loss": 1.9639, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.03332895945413988, | |
| "grad_norm": 1.3149683475494385, | |
| "learning_rate": 5.196299078795344e-05, | |
| "loss": 1.9173, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.033460175829943574, | |
| "grad_norm": 1.3609029054641724, | |
| "learning_rate": 5.1635954141088813e-05, | |
| "loss": 2.0152, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.03359139220574728, | |
| "grad_norm": 1.6170399188995361, | |
| "learning_rate": 5.1308847415393666e-05, | |
| "loss": 2.0613, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.033722608581550975, | |
| "grad_norm": 2.1593005657196045, | |
| "learning_rate": 5.0981684623031415e-05, | |
| "loss": 2.0864, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.03385382495735468, | |
| "grad_norm": 2.1808066368103027, | |
| "learning_rate": 5.0654479778567223e-05, | |
| "loss": 2.0672, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.033985041333158376, | |
| "grad_norm": 2.3150367736816406, | |
| "learning_rate": 5.0327246898367597e-05, | |
| "loss": 2.0188, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.03411625770896208, | |
| "grad_norm": 2.590977430343628, | |
| "learning_rate": 5e-05, | |
| "loss": 2.1361, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.03424747408476578, | |
| "grad_norm": 2.5015792846679688, | |
| "learning_rate": 4.9672753101632415e-05, | |
| "loss": 1.9808, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.03437869046056948, | |
| "grad_norm": 3.2308530807495117, | |
| "learning_rate": 4.934552022143279e-05, | |
| "loss": 2.0869, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.03450990683637318, | |
| "grad_norm": 3.2150726318359375, | |
| "learning_rate": 4.901831537696859e-05, | |
| "loss": 1.9293, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.03464112321217688, | |
| "grad_norm": 2.861090898513794, | |
| "learning_rate": 4.869115258460635e-05, | |
| "loss": 2.0099, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.03477233958798058, | |
| "grad_norm": 2.8258917331695557, | |
| "learning_rate": 4.83640458589112e-05, | |
| "loss": 1.8759, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.03490355596378428, | |
| "grad_norm": 2.8827760219573975, | |
| "learning_rate": 4.8037009212046586e-05, | |
| "loss": 1.8166, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.03503477233958798, | |
| "grad_norm": 2.917901039123535, | |
| "learning_rate": 4.7710056653173976e-05, | |
| "loss": 1.8738, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.03516598871539168, | |
| "grad_norm": 3.3789291381835938, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 1.8185, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.03529720509119538, | |
| "grad_norm": 3.0679268836975098, | |
| "learning_rate": 4.7056459817440544e-05, | |
| "loss": 1.7584, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.03542842146699908, | |
| "grad_norm": 3.146171808242798, | |
| "learning_rate": 4.6729843538492847e-05, | |
| "loss": 1.7729, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.035559637842802784, | |
| "grad_norm": 3.02197265625, | |
| "learning_rate": 4.640336734216403e-05, | |
| "loss": 1.7293, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.03569085421860648, | |
| "grad_norm": 3.030005931854248, | |
| "learning_rate": 4.607704521360776e-05, | |
| "loss": 1.7331, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.035822070594410185, | |
| "grad_norm": 2.7871899604797363, | |
| "learning_rate": 4.575089113137792e-05, | |
| "loss": 1.5662, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.03595328697021388, | |
| "grad_norm": 3.5039308071136475, | |
| "learning_rate": 4.542491906682989e-05, | |
| "loss": 1.7353, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.03608450334601758, | |
| "grad_norm": 3.1538591384887695, | |
| "learning_rate": 4.509914298352197e-05, | |
| "loss": 1.5676, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.036215719721821284, | |
| "grad_norm": 2.9758079051971436, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.6357, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.03634693609762498, | |
| "grad_norm": 3.1648526191711426, | |
| "learning_rate": 4.444823457228612e-05, | |
| "loss": 1.5309, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.036478152473428685, | |
| "grad_norm": 3.3066608905792236, | |
| "learning_rate": 4.412313012710813e-05, | |
| "loss": 1.572, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.03660936884923238, | |
| "grad_norm": 3.489302635192871, | |
| "learning_rate": 4.379827742747575e-05, | |
| "loss": 1.6067, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.036740585225036086, | |
| "grad_norm": 3.4767446517944336, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 1.6319, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.03687180160083978, | |
| "grad_norm": 3.145461320877075, | |
| "learning_rate": 4.3149382915901606e-05, | |
| "loss": 1.4674, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.03700301797664349, | |
| "grad_norm": 3.5818727016448975, | |
| "learning_rate": 4.282536890044104e-05, | |
| "loss": 1.566, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.037134234352447185, | |
| "grad_norm": 3.861572027206421, | |
| "learning_rate": 4.250166222229774e-05, | |
| "loss": 1.5454, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.03726545072825089, | |
| "grad_norm": 3.507399320602417, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 1.3924, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.037396667104054586, | |
| "grad_norm": 3.6404268741607666, | |
| "learning_rate": 4.185522633027057e-05, | |
| "loss": 1.4609, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.03752788347985828, | |
| "grad_norm": 3.455463171005249, | |
| "learning_rate": 4.153252480754877e-05, | |
| "loss": 1.293, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.03765909985566199, | |
| "grad_norm": 3.82511305809021, | |
| "learning_rate": 4.1210186003282275e-05, | |
| "loss": 1.3777, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.037790316231465684, | |
| "grad_norm": 4.107599258422852, | |
| "learning_rate": 4.088822372539263e-05, | |
| "loss": 1.4122, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.03792153260726939, | |
| "grad_norm": 3.8363401889801025, | |
| "learning_rate": 4.0566651765672246e-05, | |
| "loss": 1.1753, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.038052748983073086, | |
| "grad_norm": 3.759277105331421, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 1.2129, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.03818396535887679, | |
| "grad_norm": 3.4467883110046387, | |
| "learning_rate": 3.992473388371915e-05, | |
| "loss": 1.1519, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.03831518173468049, | |
| "grad_norm": 3.933537483215332, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 1.148, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.03844639811048419, | |
| "grad_norm": 3.5908944606781006, | |
| "learning_rate": 3.928454234674747e-05, | |
| "loss": 1.0006, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.03857761448628789, | |
| "grad_norm": 3.9158682823181152, | |
| "learning_rate": 3.896512824892495e-05, | |
| "loss": 1.1491, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.03870883086209159, | |
| "grad_norm": 3.2645325660705566, | |
| "learning_rate": 3.864618684828134e-05, | |
| "loss": 0.6771, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.03884004723789529, | |
| "grad_norm": 3.201977252960205, | |
| "learning_rate": 3.832773180720475e-05, | |
| "loss": 0.6873, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.038971263613698987, | |
| "grad_norm": 3.315368175506592, | |
| "learning_rate": 3.800977676724919e-05, | |
| "loss": 0.6829, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.03910247998950269, | |
| "grad_norm": 3.72304368019104, | |
| "learning_rate": 3.769233534855035e-05, | |
| "loss": 0.8417, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.03923369636530639, | |
| "grad_norm": 3.146021842956543, | |
| "learning_rate": 3.73754211492421e-05, | |
| "loss": 0.5809, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.03936491274111009, | |
| "grad_norm": 3.441509485244751, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 0.5907, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03936491274111009, | |
| "eval_loss": 1.6083000898361206, | |
| "eval_runtime": 1364.8957, | |
| "eval_samples_per_second": 9.404, | |
| "eval_steps_per_second": 2.351, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03949612911691379, | |
| "grad_norm": 1.1431409120559692, | |
| "learning_rate": 3.6743228687829595e-05, | |
| "loss": 1.9111, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.03962734549271749, | |
| "grad_norm": 1.1873606443405151, | |
| "learning_rate": 3.642797750674629e-05, | |
| "loss": 1.8401, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.03975856186852119, | |
| "grad_norm": 1.1036310195922852, | |
| "learning_rate": 3.6113307705935396e-05, | |
| "loss": 1.9566, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.039889778244324894, | |
| "grad_norm": 1.2092435359954834, | |
| "learning_rate": 3.579923276480387e-05, | |
| "loss": 1.9171, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.04002099462012859, | |
| "grad_norm": 1.284285545349121, | |
| "learning_rate": 3.5485766137276894e-05, | |
| "loss": 1.8699, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.040152210995932296, | |
| "grad_norm": 1.4310280084609985, | |
| "learning_rate": 3.5172921251221455e-05, | |
| "loss": 1.9898, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.04028342737173599, | |
| "grad_norm": 1.8922241926193237, | |
| "learning_rate": 3.486071150787128e-05, | |
| "loss": 2.0229, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.04041464374753969, | |
| "grad_norm": 2.3455190658569336, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.1236, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.040545860123343394, | |
| "grad_norm": 2.310145616531372, | |
| "learning_rate": 3.423825091761153e-05, | |
| "loss": 2.0076, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.04067707649914709, | |
| "grad_norm": 2.770493507385254, | |
| "learning_rate": 3.392802673484193e-05, | |
| "loss": 2.1042, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.040808292874950795, | |
| "grad_norm": 2.8514256477355957, | |
| "learning_rate": 3.361849102191533e-05, | |
| "loss": 2.1725, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.04093950925075449, | |
| "grad_norm": 2.9664652347564697, | |
| "learning_rate": 3.330965703831146e-05, | |
| "loss": 1.9951, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.0410707256265582, | |
| "grad_norm": 3.0636472702026367, | |
| "learning_rate": 3.300153801345028e-05, | |
| "loss": 2.0478, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.041201942002361894, | |
| "grad_norm": 3.0987637042999268, | |
| "learning_rate": 3.2694147146125345e-05, | |
| "loss": 1.8762, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.0413331583781656, | |
| "grad_norm": 3.3828492164611816, | |
| "learning_rate": 3.2387497603938326e-05, | |
| "loss": 2.0024, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.041464374753969295, | |
| "grad_norm": 3.3837296962738037, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 1.8804, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.041595591129773, | |
| "grad_norm": 3.1520280838012695, | |
| "learning_rate": 3.177647500604252e-05, | |
| "loss": 1.7544, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.041726807505576696, | |
| "grad_norm": 3.064986228942871, | |
| "learning_rate": 3.147212812450819e-05, | |
| "loss": 1.7702, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.04185802388138039, | |
| "grad_norm": 2.9096457958221436, | |
| "learning_rate": 3.116857491533947e-05, | |
| "loss": 1.7361, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.0419892402571841, | |
| "grad_norm": 3.204127073287964, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 2.0178, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.042120456632987795, | |
| "grad_norm": 3.014491081237793, | |
| "learning_rate": 3.056390149238022e-05, | |
| "loss": 1.6361, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.0422516730087915, | |
| "grad_norm": 2.9719576835632324, | |
| "learning_rate": 3.0262807180786647e-05, | |
| "loss": 1.5053, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.042382889384595196, | |
| "grad_norm": 2.9963722229003906, | |
| "learning_rate": 2.996255834484296e-05, | |
| "loss": 1.6507, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.0425141057603989, | |
| "grad_norm": 3.126011848449707, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 1.6596, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.0426453221362026, | |
| "grad_norm": 3.2865161895751953, | |
| "learning_rate": 2.936464850978027e-05, | |
| "loss": 1.742, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.0427765385120063, | |
| "grad_norm": 3.1985416412353516, | |
| "learning_rate": 2.9067013123128613e-05, | |
| "loss": 1.592, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.04290775488781, | |
| "grad_norm": 3.0526158809661865, | |
| "learning_rate": 2.8770274435964355e-05, | |
| "loss": 1.6197, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.043038971263613696, | |
| "grad_norm": 3.0706164836883545, | |
| "learning_rate": 2.8474445159585235e-05, | |
| "loss": 1.6091, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.0431701876394174, | |
| "grad_norm": 3.1661648750305176, | |
| "learning_rate": 2.8179537966332887e-05, | |
| "loss": 1.4687, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.0433014040152211, | |
| "grad_norm": 3.272674798965454, | |
| "learning_rate": 2.7885565489049946e-05, | |
| "loss": 1.5756, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.0434326203910248, | |
| "grad_norm": 3.0554146766662598, | |
| "learning_rate": 2.759254032053888e-05, | |
| "loss": 1.3903, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.0435638367668285, | |
| "grad_norm": 3.462747097015381, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 1.433, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.0436950531426322, | |
| "grad_norm": 3.4860434532165527, | |
| "learning_rate": 2.700938207760701e-05, | |
| "loss": 1.4024, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.0438262695184359, | |
| "grad_norm": 3.317476987838745, | |
| "learning_rate": 2.671927398374443e-05, | |
| "loss": 1.3322, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.043957485894239604, | |
| "grad_norm": 3.5378825664520264, | |
| "learning_rate": 2.6430163158700115e-05, | |
| "loss": 1.3707, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.0440887022700433, | |
| "grad_norm": 3.7148430347442627, | |
| "learning_rate": 2.6142061987019577e-05, | |
| "loss": 1.4425, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.044219918645847005, | |
| "grad_norm": 3.060731887817383, | |
| "learning_rate": 2.5854982809998153e-05, | |
| "loss": 1.1957, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.0443511350216507, | |
| "grad_norm": 3.4139750003814697, | |
| "learning_rate": 2.556893792515227e-05, | |
| "loss": 1.3196, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.0444823513974544, | |
| "grad_norm": 3.1842236518859863, | |
| "learning_rate": 2.5283939585692783e-05, | |
| "loss": 1.182, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.0446135677732581, | |
| "grad_norm": 3.5313189029693604, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.1865, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.0447447841490618, | |
| "grad_norm": 3.486128091812134, | |
| "learning_rate": 2.471713133110078e-05, | |
| "loss": 1.1062, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.044876000524865504, | |
| "grad_norm": 3.0605080127716064, | |
| "learning_rate": 2.4435345696147403e-05, | |
| "loss": 0.8957, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.0450072169006692, | |
| "grad_norm": 3.2468960285186768, | |
| "learning_rate": 2.4154655165898627e-05, | |
| "loss": 0.9158, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.045138433276472906, | |
| "grad_norm": 3.440025568008423, | |
| "learning_rate": 2.3875071764202563e-05, | |
| "loss": 0.8078, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.0452696496522766, | |
| "grad_norm": 2.9829273223876953, | |
| "learning_rate": 2.3596607467481603e-05, | |
| "loss": 0.7363, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.04540086602808031, | |
| "grad_norm": 3.14467453956604, | |
| "learning_rate": 2.3319274204219428e-05, | |
| "loss": 0.5925, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.045532082403884004, | |
| "grad_norm": 3.064061403274536, | |
| "learning_rate": 2.3043083854449988e-05, | |
| "loss": 0.6579, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.04566329877968771, | |
| "grad_norm": 2.8710074424743652, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 0.5748, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.045794515155491405, | |
| "grad_norm": 3.819432020187378, | |
| "learning_rate": 2.2494179170225333e-05, | |
| "loss": 0.6714, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.0459257315312951, | |
| "grad_norm": 4.986090183258057, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.7477, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.04605694790709881, | |
| "grad_norm": 2.6902003288269043, | |
| "learning_rate": 2.194998746679952e-05, | |
| "loss": 1.8673, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.046188164282902504, | |
| "grad_norm": 2.166841506958008, | |
| "learning_rate": 2.167968815375837e-05, | |
| "loss": 1.8827, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.04631938065870621, | |
| "grad_norm": 1.7183693647384644, | |
| "learning_rate": 2.1410601988619394e-05, | |
| "loss": 1.8546, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.046450597034509905, | |
| "grad_norm": 1.265217900276184, | |
| "learning_rate": 2.1142740498138324e-05, | |
| "loss": 1.9024, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.04658181341031361, | |
| "grad_norm": 1.721006155014038, | |
| "learning_rate": 2.08761151566099e-05, | |
| "loss": 1.955, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.046713029786117306, | |
| "grad_norm": 2.1906604766845703, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.01, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.04684424616192101, | |
| "grad_norm": 2.1465506553649902, | |
| "learning_rate": 2.034661855233815e-05, | |
| "loss": 2.0038, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.04697546253772471, | |
| "grad_norm": 2.263490915298462, | |
| "learning_rate": 2.008376997146705e-05, | |
| "loss": 2.068, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.04710667891352841, | |
| "grad_norm": 2.7283573150634766, | |
| "learning_rate": 1.982220290232143e-05, | |
| "loss": 2.0493, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.04723789528933211, | |
| "grad_norm": 2.6310789585113525, | |
| "learning_rate": 1.9561928549563968e-05, | |
| "loss": 2.069, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.047369111665135806, | |
| "grad_norm": 2.767486333847046, | |
| "learning_rate": 1.9302958062481673e-05, | |
| "loss": 2.028, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.04750032804093951, | |
| "grad_norm": 2.8701977729797363, | |
| "learning_rate": 1.9045302534508297e-05, | |
| "loss": 1.9971, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.04763154441674321, | |
| "grad_norm": 3.023191213607788, | |
| "learning_rate": 1.8788973002749112e-05, | |
| "loss": 1.9896, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.04776276079254691, | |
| "grad_norm": 3.0075771808624268, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 1.9956, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.04789397716835061, | |
| "grad_norm": 3.155802011489868, | |
| "learning_rate": 1.8280335791817733e-05, | |
| "loss": 1.7729, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.04802519354415431, | |
| "grad_norm": 3.0290050506591797, | |
| "learning_rate": 1.8028049900970767e-05, | |
| "loss": 1.7798, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.04815640991995801, | |
| "grad_norm": 2.8980050086975098, | |
| "learning_rate": 1.777713358205514e-05, | |
| "loss": 1.7175, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.048287626295761714, | |
| "grad_norm": 3.2710909843444824, | |
| "learning_rate": 1.7527597583490822e-05, | |
| "loss": 1.829, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.04841884267156541, | |
| "grad_norm": 3.313262462615967, | |
| "learning_rate": 1.7279452594569483e-05, | |
| "loss": 1.8382, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.04855005904736911, | |
| "grad_norm": 3.1630051136016846, | |
| "learning_rate": 1.703270924499656e-05, | |
| "loss": 1.7057, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.04868127542317281, | |
| "grad_norm": 3.111182928085327, | |
| "learning_rate": 1.678737810443593e-05, | |
| "loss": 1.6428, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.04881249179897651, | |
| "grad_norm": 3.932502508163452, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 1.7103, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.048943708174780214, | |
| "grad_norm": 3.1841988563537598, | |
| "learning_rate": 1.6300994426085103e-05, | |
| "loss": 1.5631, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.04907492455058391, | |
| "grad_norm": 3.1388206481933594, | |
| "learning_rate": 1.605996272335291e-05, | |
| "loss": 1.6667, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.049206140926387615, | |
| "grad_norm": 3.2520394325256348, | |
| "learning_rate": 1.5820384898856434e-05, | |
| "loss": 1.5319, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.04933735730219131, | |
| "grad_norm": 3.448615074157715, | |
| "learning_rate": 1.5582271215312294e-05, | |
| "loss": 1.3435, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.049468573677995016, | |
| "grad_norm": 3.1773369312286377, | |
| "learning_rate": 1.5345631872718214e-05, | |
| "loss": 1.4977, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.04959979005379871, | |
| "grad_norm": 3.773745536804199, | |
| "learning_rate": 1.5110477007916001e-05, | |
| "loss": 1.6206, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.04973100642960242, | |
| "grad_norm": 3.2866322994232178, | |
| "learning_rate": 1.4876816694157419e-05, | |
| "loss": 1.4102, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.049862222805406115, | |
| "grad_norm": 3.218993902206421, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 1.4885, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.04999343918120981, | |
| "grad_norm": 3.413767099380493, | |
| "learning_rate": 1.4414019692241437e-05, | |
| "loss": 1.4064, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.050124655557013516, | |
| "grad_norm": 3.1845431327819824, | |
| "learning_rate": 1.4184902828767287e-05, | |
| "loss": 1.3509, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.05025587193281721, | |
| "grad_norm": 3.4849841594696045, | |
| "learning_rate": 1.3957320164854059e-05, | |
| "loss": 1.4778, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.05038708830862092, | |
| "grad_norm": 3.433154344558716, | |
| "learning_rate": 1.373128144938563e-05, | |
| "loss": 1.2556, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.050518304684424614, | |
| "grad_norm": 3.4509999752044678, | |
| "learning_rate": 1.3506796365108232e-05, | |
| "loss": 1.3327, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.05064952106022832, | |
| "grad_norm": 3.3174984455108643, | |
| "learning_rate": 1.3283874528215733e-05, | |
| "loss": 1.2881, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.050780737436032015, | |
| "grad_norm": 3.912639856338501, | |
| "learning_rate": 1.3062525487937699e-05, | |
| "loss": 1.2161, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.05091195381183572, | |
| "grad_norm": 3.7738358974456787, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.4717, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.05104317018763942, | |
| "grad_norm": 3.3311028480529785, | |
| "learning_rate": 1.2624583656870154e-05, | |
| "loss": 0.9475, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.05117438656344312, | |
| "grad_norm": 3.2263801097869873, | |
| "learning_rate": 1.2408009626051137e-05, | |
| "loss": 1.0097, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.05130560293924682, | |
| "grad_norm": 3.5462255477905273, | |
| "learning_rate": 1.2193045910983863e-05, | |
| "loss": 1.0301, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.051436819315050515, | |
| "grad_norm": 3.075239419937134, | |
| "learning_rate": 1.1979701719998453e-05, | |
| "loss": 0.8822, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.05156803569085422, | |
| "grad_norm": 3.8291819095611572, | |
| "learning_rate": 1.1767986192049984e-05, | |
| "loss": 1.0269, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.051699252066657916, | |
| "grad_norm": 3.1441256999969482, | |
| "learning_rate": 1.1557908396327028e-05, | |
| "loss": 0.8817, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.05183046844246162, | |
| "grad_norm": 3.654750347137451, | |
| "learning_rate": 1.134947733186315e-05, | |
| "loss": 0.8977, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.05196168481826532, | |
| "grad_norm": 3.509984254837036, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 0.7434, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.05209290119406902, | |
| "grad_norm": 3.4480764865875244, | |
| "learning_rate": 1.0937591039762085e-05, | |
| "loss": 0.6737, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.05222411756987272, | |
| "grad_norm": 3.189035415649414, | |
| "learning_rate": 1.0734153455962765e-05, | |
| "loss": 0.5615, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.05235533394567642, | |
| "grad_norm": 2.6568944454193115, | |
| "learning_rate": 1.0532397890342505e-05, | |
| "loss": 0.4878, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.05248655032148012, | |
| "grad_norm": 4.114481449127197, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 0.6082, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.05248655032148012, | |
| "eval_loss": 1.4705314636230469, | |
| "eval_runtime": 1366.8691, | |
| "eval_samples_per_second": 9.391, | |
| "eval_steps_per_second": 2.348, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.052617766697283824, | |
| "grad_norm": 1.5547120571136475, | |
| "learning_rate": 1.013396731136465e-05, | |
| "loss": 1.8584, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.05274898307308752, | |
| "grad_norm": 1.4148300886154175, | |
| "learning_rate": 9.937309365446973e-06, | |
| "loss": 1.8086, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.05288019944889122, | |
| "grad_norm": 1.1794555187225342, | |
| "learning_rate": 9.742367571857091e-06, | |
| "loss": 1.7795, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.05301141582469492, | |
| "grad_norm": 1.2620103359222412, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.8577, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.05314263220049862, | |
| "grad_norm": 1.2999380826950073, | |
| "learning_rate": 9.357665770419244e-06, | |
| "loss": 1.9158, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.053273848576302324, | |
| "grad_norm": 1.7266931533813477, | |
| "learning_rate": 9.167922241916055e-06, | |
| "loss": 1.9962, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.05340506495210602, | |
| "grad_norm": 1.747637152671814, | |
| "learning_rate": 8.97992782372432e-06, | |
| "loss": 1.9808, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.053536281327909725, | |
| "grad_norm": 2.2035903930664062, | |
| "learning_rate": 8.793690568899216e-06, | |
| "loss": 1.975, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.05366749770371342, | |
| "grad_norm": 2.1186153888702393, | |
| "learning_rate": 8.609218455224893e-06, | |
| "loss": 1.9926, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.053798714079517126, | |
| "grad_norm": 2.213442087173462, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 1.9039, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.053929930455320824, | |
| "grad_norm": 2.3685803413391113, | |
| "learning_rate": 8.245601184062852e-06, | |
| "loss": 1.958, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.05406114683112453, | |
| "grad_norm": 2.664763927459717, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 1.9546, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.054192363206928225, | |
| "grad_norm": 2.719207286834717, | |
| "learning_rate": 7.889138314185678e-06, | |
| "loss": 1.9461, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.05432357958273192, | |
| "grad_norm": 2.6173901557922363, | |
| "learning_rate": 7.71360891480134e-06, | |
| "loss": 1.8008, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.054454795958535626, | |
| "grad_norm": 3.068253755569458, | |
| "learning_rate": 7.539890923671062e-06, | |
| "loss": 1.9327, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.05458601233433932, | |
| "grad_norm": 2.894899606704712, | |
| "learning_rate": 7.367991782295391e-06, | |
| "loss": 1.9416, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.05471722871014303, | |
| "grad_norm": 3.028637170791626, | |
| "learning_rate": 7.197918854261432e-06, | |
| "loss": 1.7818, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.054848445085946725, | |
| "grad_norm": 2.8751864433288574, | |
| "learning_rate": 7.029679424927365e-06, | |
| "loss": 1.8194, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.05497966146175043, | |
| "grad_norm": 3.282900810241699, | |
| "learning_rate": 6.863280701110408e-06, | |
| "loss": 1.8425, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.055110877837554126, | |
| "grad_norm": 3.1004960536956787, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 1.6386, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.05524209421335783, | |
| "grad_norm": 3.0315091609954834, | |
| "learning_rate": 6.536033802742813e-06, | |
| "loss": 1.792, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.05537331058916153, | |
| "grad_norm": 3.5640015602111816, | |
| "learning_rate": 6.375199646360142e-06, | |
| "loss": 1.8219, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.055504526964965224, | |
| "grad_norm": 3.106428861618042, | |
| "learning_rate": 6.216234231230012e-06, | |
| "loss": 1.5988, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.05563574334076893, | |
| "grad_norm": 3.2876532077789307, | |
| "learning_rate": 6.059144366901736e-06, | |
| "loss": 1.6089, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.055766959716572626, | |
| "grad_norm": 3.1995625495910645, | |
| "learning_rate": 5.903936782582253e-06, | |
| "loss": 1.597, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.05589817609237633, | |
| "grad_norm": 3.096595287322998, | |
| "learning_rate": 5.750618126847912e-06, | |
| "loss": 1.3686, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.05602939246818003, | |
| "grad_norm": 3.0396366119384766, | |
| "learning_rate": 5.599194967359639e-06, | |
| "loss": 1.5684, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.05616060884398373, | |
| "grad_norm": 3.076733350753784, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 1.4296, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.05629182521978743, | |
| "grad_norm": 3.201915740966797, | |
| "learning_rate": 5.302061001503394e-06, | |
| "loss": 1.6379, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.05642304159559113, | |
| "grad_norm": 3.393874168395996, | |
| "learning_rate": 5.156362923365588e-06, | |
| "loss": 1.5963, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.05655425797139483, | |
| "grad_norm": 3.4268441200256348, | |
| "learning_rate": 5.012585797388936e-06, | |
| "loss": 1.3698, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.05668547434719853, | |
| "grad_norm": 3.298431158065796, | |
| "learning_rate": 4.87073578250698e-06, | |
| "loss": 1.5274, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.05681669072300223, | |
| "grad_norm": 3.310758352279663, | |
| "learning_rate": 4.730818955102234e-06, | |
| "loss": 1.2892, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.05694790709880593, | |
| "grad_norm": 3.1434218883514404, | |
| "learning_rate": 4.592841308745932e-06, | |
| "loss": 1.2763, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.05707912347460963, | |
| "grad_norm": 3.1354024410247803, | |
| "learning_rate": 4.456808753941205e-06, | |
| "loss": 1.2465, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.05721033985041333, | |
| "grad_norm": 3.235828161239624, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 1.1989, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.05734155622621703, | |
| "grad_norm": 3.4798684120178223, | |
| "learning_rate": 4.190602144143207e-06, | |
| "loss": 1.3004, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.05747277260202073, | |
| "grad_norm": 3.5043392181396484, | |
| "learning_rate": 4.06043949255509e-06, | |
| "loss": 1.1935, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.057603988977824434, | |
| "grad_norm": 3.677044630050659, | |
| "learning_rate": 3.932244738840379e-06, | |
| "loss": 1.2309, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.05773520535362813, | |
| "grad_norm": 4.231906414031982, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 1.2004, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.057866421729431836, | |
| "grad_norm": 3.8758881092071533, | |
| "learning_rate": 3.681780806244095e-06, | |
| "loss": 1.1777, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.05799763810523553, | |
| "grad_norm": 3.5995683670043945, | |
| "learning_rate": 3.5595223564037884e-06, | |
| "loss": 1.1675, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.05812885448103924, | |
| "grad_norm": 3.3509786128997803, | |
| "learning_rate": 3.4392532620598216e-06, | |
| "loss": 1.1342, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.058260070856842934, | |
| "grad_norm": 3.3423001766204834, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 1.0987, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.05839128723264663, | |
| "grad_norm": 3.2946929931640625, | |
| "learning_rate": 3.2047036621337236e-06, | |
| "loss": 0.8206, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.058522503608450335, | |
| "grad_norm": 3.773805618286133, | |
| "learning_rate": 3.0904332038757977e-06, | |
| "loss": 0.792, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.05865371998425403, | |
| "grad_norm": 3.8902366161346436, | |
| "learning_rate": 2.978172195332263e-06, | |
| "loss": 0.9764, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.05878493636005774, | |
| "grad_norm": 3.055663824081421, | |
| "learning_rate": 2.8679254453910785e-06, | |
| "loss": 0.5963, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.058916152735861434, | |
| "grad_norm": 2.5872833728790283, | |
| "learning_rate": 2.759697676656098e-06, | |
| "loss": 0.4251, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.05904736911166514, | |
| "grad_norm": 2.976865530014038, | |
| "learning_rate": 2.653493525244721e-06, | |
| "loss": 0.5668, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.059178585487468835, | |
| "grad_norm": 1.0085222721099854, | |
| "learning_rate": 2.549317540589308e-06, | |
| "loss": 1.8009, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.05930980186327254, | |
| "grad_norm": 1.0669842958450317, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.7289, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.059441018239076236, | |
| "grad_norm": 1.1122716665267944, | |
| "learning_rate": 2.3470678346851518e-06, | |
| "loss": 1.8283, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.05957223461487994, | |
| "grad_norm": 1.1549385786056519, | |
| "learning_rate": 2.2490027771406687e-06, | |
| "loss": 1.925, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.05970345099068364, | |
| "grad_norm": 1.1655536890029907, | |
| "learning_rate": 2.152983213389559e-06, | |
| "loss": 1.9207, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.059834667366487335, | |
| "grad_norm": 1.4617455005645752, | |
| "learning_rate": 2.0590132565903476e-06, | |
| "loss": 1.9206, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.05996588374229104, | |
| "grad_norm": 1.655059576034546, | |
| "learning_rate": 1.9670969321032407e-06, | |
| "loss": 1.9933, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.060097100118094736, | |
| "grad_norm": 1.801466703414917, | |
| "learning_rate": 1.8772381773176417e-06, | |
| "loss": 1.9565, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.06022831649389844, | |
| "grad_norm": 1.9467339515686035, | |
| "learning_rate": 1.7894408414835362e-06, | |
| "loss": 1.9179, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.06035953286970214, | |
| "grad_norm": 2.1917128562927246, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 1.9006, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.06049074924550584, | |
| "grad_norm": 2.31794810295105, | |
| "learning_rate": 1.620045381987012e-06, | |
| "loss": 1.8937, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.06062196562130954, | |
| "grad_norm": 2.556521415710449, | |
| "learning_rate": 1.5384545146622852e-06, | |
| "loss": 1.9225, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.06075318199711324, | |
| "grad_norm": 2.7441983222961426, | |
| "learning_rate": 1.4589395786535953e-06, | |
| "loss": 1.7219, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.06088439837291694, | |
| "grad_norm": 2.604498863220215, | |
| "learning_rate": 1.3815039801161721e-06, | |
| "loss": 1.7661, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.061015614748720644, | |
| "grad_norm": 2.8174169063568115, | |
| "learning_rate": 1.3061510361333185e-06, | |
| "loss": 1.8408, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.06114683112452434, | |
| "grad_norm": 2.8200817108154297, | |
| "learning_rate": 1.232883974574367e-06, | |
| "loss": 1.7759, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.06127804750032804, | |
| "grad_norm": 3.005772829055786, | |
| "learning_rate": 1.1617059339563807e-06, | |
| "loss": 1.7675, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.06140926387613174, | |
| "grad_norm": 2.790365219116211, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 1.6673, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.06154048025193544, | |
| "grad_norm": 2.7628681659698486, | |
| "learning_rate": 1.0256290220474307e-06, | |
| "loss": 1.5659, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.06167169662773914, | |
| "grad_norm": 2.9881839752197266, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 1.6035, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.06180291300354284, | |
| "grad_norm": 3.119401693344116, | |
| "learning_rate": 8.979436164848088e-07, | |
| "loss": 1.7385, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.061934129379346545, | |
| "grad_norm": 3.145608425140381, | |
| "learning_rate": 8.372546218022747e-07, | |
| "loss": 1.8044, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.06206534575515024, | |
| "grad_norm": 3.2894794940948486, | |
| "learning_rate": 7.786715955054203e-07, | |
| "loss": 1.6499, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.062196562130953946, | |
| "grad_norm": 3.073772430419922, | |
| "learning_rate": 7.221970470961125e-07, | |
| "loss": 1.6461, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.06232777850675764, | |
| "grad_norm": 2.914522647857666, | |
| "learning_rate": 6.678333957560512e-07, | |
| "loss": 1.4814, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.06245899488256134, | |
| "grad_norm": 3.2599356174468994, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 1.5119, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.06259021125836504, | |
| "grad_norm": 3.0420191287994385, | |
| "learning_rate": 5.654480087916303e-07, | |
| "loss": 1.5962, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.06272142763416874, | |
| "grad_norm": 3.278502941131592, | |
| "learning_rate": 5.174306590164879e-07, | |
| "loss": 1.5504, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.06285264400997244, | |
| "grad_norm": 2.9558990001678467, | |
| "learning_rate": 4.715329778211375e-07, | |
| "loss": 1.4017, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.06298386038577615, | |
| "grad_norm": 3.2562479972839355, | |
| "learning_rate": 4.277569313094809e-07, | |
| "loss": 1.3569, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.06311507676157985, | |
| "grad_norm": 3.3667497634887695, | |
| "learning_rate": 3.8610439470164737e-07, | |
| "loss": 1.452, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.06324629313738354, | |
| "grad_norm": 3.1160595417022705, | |
| "learning_rate": 3.465771522536854e-07, | |
| "loss": 1.2796, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.06337750951318724, | |
| "grad_norm": 3.3417913913726807, | |
| "learning_rate": 3.09176897181096e-07, | |
| "loss": 1.4638, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.06350872588899095, | |
| "grad_norm": 3.3148908615112305, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 1.3271, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.06363994226479465, | |
| "grad_norm": 3.5485448837280273, | |
| "learning_rate": 2.407636663901591e-07, | |
| "loss": 1.4274, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.06377115864059835, | |
| "grad_norm": 3.49263858795166, | |
| "learning_rate": 2.0975362126691712e-07, | |
| "loss": 1.3793, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.06390237501640204, | |
| "grad_norm": 3.290903091430664, | |
| "learning_rate": 1.8087642458373134e-07, | |
| "loss": 1.2698, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.06403359139220574, | |
| "grad_norm": 3.2131378650665283, | |
| "learning_rate": 1.5413331334360182e-07, | |
| "loss": 1.1554, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.06416480776800945, | |
| "grad_norm": 3.7000808715820312, | |
| "learning_rate": 1.2952543313240472e-07, | |
| "loss": 1.2878, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.06429602414381315, | |
| "grad_norm": 3.0949819087982178, | |
| "learning_rate": 1.0705383806982606e-07, | |
| "loss": 1.0326, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.06442724051961685, | |
| "grad_norm": 3.2864902019500732, | |
| "learning_rate": 8.671949076420882e-08, | |
| "loss": 1.041, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.06455845689542054, | |
| "grad_norm": 3.193171262741089, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 0.9282, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.06468967327122425, | |
| "grad_norm": 3.8569085597991943, | |
| "learning_rate": 5.246593205699424e-08, | |
| "loss": 0.8974, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.06482088964702795, | |
| "grad_norm": 3.4853317737579346, | |
| "learning_rate": 3.8548187963854956e-08, | |
| "loss": 0.797, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.06495210602283165, | |
| "grad_norm": 3.2885305881500244, | |
| "learning_rate": 2.6770626181715773e-08, | |
| "loss": 0.9086, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.06508332239863535, | |
| "grad_norm": 3.696621894836426, | |
| "learning_rate": 1.7133751222137007e-08, | |
| "loss": 0.8038, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.06521453877443904, | |
| "grad_norm": 3.6335575580596924, | |
| "learning_rate": 9.637975896759077e-09, | |
| "loss": 0.9278, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.06534575515024275, | |
| "grad_norm": 3.405994176864624, | |
| "learning_rate": 4.2836212996499865e-09, | |
| "loss": 0.7081, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.06547697152604645, | |
| "grad_norm": 2.7525062561035156, | |
| "learning_rate": 1.0709167935385455e-09, | |
| "loss": 0.5322, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.06560818790185015, | |
| "grad_norm": 3.2091686725616455, | |
| "learning_rate": 0.0, | |
| "loss": 0.491, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06560818790185015, | |
| "eval_loss": 1.454202651977539, | |
| "eval_runtime": 1356.7364, | |
| "eval_samples_per_second": 9.461, | |
| "eval_steps_per_second": 2.365, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.026400488001372e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |