| { | |
| "best_metric": 1.116738200187683, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 2.168021680216802, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01084010840108401, | |
| "grad_norm": 21.86941909790039, | |
| "learning_rate": 1e-05, | |
| "loss": 5.9483, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01084010840108401, | |
| "eval_loss": 2.248394012451172, | |
| "eval_runtime": 11.3589, | |
| "eval_samples_per_second": 13.734, | |
| "eval_steps_per_second": 3.433, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02168021680216802, | |
| "grad_norm": 42.31118392944336, | |
| "learning_rate": 2e-05, | |
| "loss": 7.5193, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 34.676002502441406, | |
| "learning_rate": 3e-05, | |
| "loss": 7.3278, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.04336043360433604, | |
| "grad_norm": 43.211429595947266, | |
| "learning_rate": 4e-05, | |
| "loss": 6.7843, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.05420054200542006, | |
| "grad_norm": 43.430877685546875, | |
| "learning_rate": 5e-05, | |
| "loss": 6.4338, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 23.902545928955078, | |
| "learning_rate": 6e-05, | |
| "loss": 6.2412, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07588075880758807, | |
| "grad_norm": 32.86506652832031, | |
| "learning_rate": 7e-05, | |
| "loss": 5.6165, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08672086720867209, | |
| "grad_norm": 24.925737380981445, | |
| "learning_rate": 8e-05, | |
| "loss": 6.0098, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 25.875600814819336, | |
| "learning_rate": 9e-05, | |
| "loss": 5.8786, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10840108401084012, | |
| "grad_norm": 21.96430206298828, | |
| "learning_rate": 0.0001, | |
| "loss": 5.6723, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11924119241192412, | |
| "grad_norm": 22.341049194335938, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 5.2494, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 31.55384063720703, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 5.5743, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14092140921409213, | |
| "grad_norm": 23.872188568115234, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 5.4399, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15176151761517614, | |
| "grad_norm": 26.248977661132812, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 5.8143, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 20.28582191467285, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 5.1948, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17344173441734417, | |
| "grad_norm": 23.422361373901367, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 5.5889, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1842818428184282, | |
| "grad_norm": 45.05106735229492, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 4.9165, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 25.640554428100586, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 5.4901, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20596205962059622, | |
| "grad_norm": 22.076374053955078, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 5.4603, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.21680216802168023, | |
| "grad_norm": 24.726655960083008, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 4.6989, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.22764227642276422, | |
| "grad_norm": 22.62340545654297, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 5.4016, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.23848238482384823, | |
| "grad_norm": 23.824033737182617, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 5.3024, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.24932249322493225, | |
| "grad_norm": 31.84869384765625, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 5.2125, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2601626016260163, | |
| "grad_norm": 13.845771789550781, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 5.949, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.27100271002710025, | |
| "grad_norm": 18.300031661987305, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 5.9133, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.28184281842818426, | |
| "grad_norm": 16.029516220092773, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 5.4023, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 16.750877380371094, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 5.4457, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3035230352303523, | |
| "grad_norm": 15.00551700592041, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 5.7783, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3143631436314363, | |
| "grad_norm": 18.077030181884766, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 5.3401, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3252032520325203, | |
| "grad_norm": 14.956138610839844, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 4.6812, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.33604336043360433, | |
| "grad_norm": 15.319053649902344, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 5.4935, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34688346883468835, | |
| "grad_norm": 16.896076202392578, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 4.8401, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.35772357723577236, | |
| "grad_norm": 15.099456787109375, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 4.965, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3685636856368564, | |
| "grad_norm": 15.831088066101074, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 5.2788, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3794037940379404, | |
| "grad_norm": 15.724886894226074, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 4.4665, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3902439024390244, | |
| "grad_norm": 18.4834041595459, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 4.8604, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4010840108401084, | |
| "grad_norm": 17.911102294921875, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 4.7507, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.41192411924119243, | |
| "grad_norm": 19.0974178314209, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 4.2766, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.42276422764227645, | |
| "grad_norm": 20.512279510498047, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 4.8773, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.43360433604336046, | |
| "grad_norm": 17.6418399810791, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 5.4788, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 19.53268051147461, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 4.4733, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.45528455284552843, | |
| "grad_norm": 19.109052658081055, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 5.0834, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.46612466124661245, | |
| "grad_norm": 19.814104080200195, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 5.3832, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.47696476964769646, | |
| "grad_norm": 17.042949676513672, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 4.8963, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 24.227577209472656, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 4.8807, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4986449864498645, | |
| "grad_norm": 22.67888832092285, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 4.946, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5094850948509485, | |
| "grad_norm": 10.419925689697266, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 5.762, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5203252032520326, | |
| "grad_norm": 13.869553565979004, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 5.6297, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.5311653116531165, | |
| "grad_norm": 16.4982967376709, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 5.0761, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5420054200542005, | |
| "grad_norm": 13.941553115844727, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 5.1964, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5420054200542005, | |
| "eval_loss": 1.2086718082427979, | |
| "eval_runtime": 11.6494, | |
| "eval_samples_per_second": 13.391, | |
| "eval_steps_per_second": 3.348, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5528455284552846, | |
| "grad_norm": 13.97443675994873, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 5.2024, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5636856368563685, | |
| "grad_norm": 13.956024169921875, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 4.5837, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5745257452574526, | |
| "grad_norm": 13.442913055419922, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 5.0317, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 13.219365119934082, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 4.7051, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5962059620596206, | |
| "grad_norm": 12.270064353942871, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 4.6755, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6070460704607046, | |
| "grad_norm": 14.628702163696289, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 4.5528, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6178861788617886, | |
| "grad_norm": 13.100887298583984, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 4.3857, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6287262872628726, | |
| "grad_norm": 15.415169715881348, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 4.6902, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6395663956639567, | |
| "grad_norm": 13.177145004272461, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 4.4471, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6504065040650406, | |
| "grad_norm": 15.360495567321777, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 4.4916, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6612466124661247, | |
| "grad_norm": 14.919403076171875, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 4.7263, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6720867208672087, | |
| "grad_norm": 15.36029052734375, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 4.5792, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6829268292682927, | |
| "grad_norm": 16.579078674316406, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 5.2583, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6937669376693767, | |
| "grad_norm": 17.175189971923828, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 5.0342, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7046070460704607, | |
| "grad_norm": 16.527671813964844, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 4.7562, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7154471544715447, | |
| "grad_norm": 20.473243713378906, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 4.5965, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7262872628726287, | |
| "grad_norm": 20.61736297607422, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 5.2178, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7371273712737128, | |
| "grad_norm": 18.058507919311523, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 4.5031, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7479674796747967, | |
| "grad_norm": 21.392812728881836, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 5.2819, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7588075880758808, | |
| "grad_norm": 9.471242904663086, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 5.3587, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7696476964769647, | |
| "grad_norm": 13.22653865814209, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 4.6233, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7804878048780488, | |
| "grad_norm": 10.99927043914795, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 4.386, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7913279132791328, | |
| "grad_norm": 12.9929780960083, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 4.8668, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8021680216802168, | |
| "grad_norm": 14.063130378723145, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 4.6319, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8130081300813008, | |
| "grad_norm": 11.381362915039062, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 4.706, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8238482384823849, | |
| "grad_norm": 11.88614559173584, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 4.9185, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8346883468834688, | |
| "grad_norm": 12.178693771362305, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 4.6708, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8455284552845529, | |
| "grad_norm": 13.19189453125, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 4.4111, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8563685636856369, | |
| "grad_norm": 13.036541938781738, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 3.7635, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8672086720867209, | |
| "grad_norm": 13.367840766906738, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 4.71, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8780487804878049, | |
| "grad_norm": 12.398404121398926, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 4.1636, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 15.483765602111816, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 4.8336, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8997289972899729, | |
| "grad_norm": 14.48204231262207, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 4.641, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9105691056910569, | |
| "grad_norm": 14.978960990905762, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 4.9607, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9214092140921409, | |
| "grad_norm": 13.384892463684082, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 4.419, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9322493224932249, | |
| "grad_norm": 14.83173656463623, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 4.9142, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.943089430894309, | |
| "grad_norm": 14.982903480529785, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 4.4034, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9539295392953929, | |
| "grad_norm": 15.042606353759766, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 4.891, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.964769647696477, | |
| "grad_norm": 17.68012237548828, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 4.4156, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.975609756097561, | |
| "grad_norm": 16.286632537841797, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 4.5538, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.986449864498645, | |
| "grad_norm": 15.605447769165039, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 3.9257, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.997289972899729, | |
| "grad_norm": 16.65151023864746, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 4.0481, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.008130081300813, | |
| "grad_norm": 30.399702072143555, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 3.4919, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.018970189701897, | |
| "grad_norm": 7.537675857543945, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 3.5223, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.029810298102981, | |
| "grad_norm": 7.987434387207031, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 2.8859, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.040650406504065, | |
| "grad_norm": 8.36559009552002, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 3.0624, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.051490514905149, | |
| "grad_norm": 11.518872261047363, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 2.9636, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.062330623306233, | |
| "grad_norm": 9.825814247131348, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 2.7065, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.0731707317073171, | |
| "grad_norm": 10.017013549804688, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 2.7139, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.084010840108401, | |
| "grad_norm": 8.91040325164795, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 2.3674, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.084010840108401, | |
| "eval_loss": 1.145018219947815, | |
| "eval_runtime": 11.6503, | |
| "eval_samples_per_second": 13.39, | |
| "eval_steps_per_second": 3.348, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.094850948509485, | |
| "grad_norm": 10.517824172973633, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 2.3814, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.1056910569105691, | |
| "grad_norm": 11.603822708129883, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 2.1762, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1165311653116532, | |
| "grad_norm": 9.575349807739258, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 2.0891, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.127371273712737, | |
| "grad_norm": 10.596747398376465, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 2.3945, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1382113821138211, | |
| "grad_norm": 9.49907112121582, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9272, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1490514905149052, | |
| "grad_norm": 12.568108558654785, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 2.3345, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.1598915989159893, | |
| "grad_norm": 12.349029541015625, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 2.0959, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.170731707317073, | |
| "grad_norm": 9.691751480102539, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 1.8334, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.1815718157181572, | |
| "grad_norm": 13.008195877075195, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 2.3569, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.1924119241192412, | |
| "grad_norm": 16.772262573242188, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 2.9454, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.203252032520325, | |
| "grad_norm": 12.214621543884277, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 2.3502, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2140921409214092, | |
| "grad_norm": 14.069916725158691, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 2.0635, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2249322493224932, | |
| "grad_norm": 15.881183624267578, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 2.3121, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2357723577235773, | |
| "grad_norm": 11.533042907714844, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 2.065, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2466124661246614, | |
| "grad_norm": 14.450631141662598, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 2.4597, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2574525745257452, | |
| "grad_norm": 10.32443618774414, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 3.9081, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.2682926829268293, | |
| "grad_norm": 14.189412117004395, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 3.2457, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.2791327913279134, | |
| "grad_norm": 10.970619201660156, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 3.0248, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.2899728997289972, | |
| "grad_norm": 11.229940414428711, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 3.0206, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.3008130081300813, | |
| "grad_norm": 10.457233428955078, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 2.6235, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3116531165311653, | |
| "grad_norm": 12.614952087402344, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 2.8028, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3224932249322494, | |
| "grad_norm": 11.393104553222656, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 2.895, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 14.456162452697754, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 2.3521, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3441734417344173, | |
| "grad_norm": 10.501736640930176, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.5941, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.3550135501355014, | |
| "grad_norm": 10.971872329711914, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 2.2562, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.3658536585365852, | |
| "grad_norm": 9.223410606384277, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 2.2514, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.3766937669376693, | |
| "grad_norm": 10.584888458251953, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 2.0338, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.3875338753387534, | |
| "grad_norm": 11.694501876831055, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 2.3947, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.3983739837398375, | |
| "grad_norm": 9.978876113891602, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 2.1236, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4092140921409215, | |
| "grad_norm": 10.213749885559082, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 2.2082, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4200542005420054, | |
| "grad_norm": 10.720746994018555, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 1.9924, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4308943089430894, | |
| "grad_norm": 12.569414138793945, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 2.365, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.4417344173441735, | |
| "grad_norm": 11.3380765914917, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 2.0934, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.4525745257452574, | |
| "grad_norm": 11.89591121673584, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 2.156, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.4634146341463414, | |
| "grad_norm": 12.46812915802002, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 2.0201, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.4742547425474255, | |
| "grad_norm": 11.85822582244873, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 2.2182, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.4850948509485096, | |
| "grad_norm": 12.369199752807617, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 1.8484, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.4959349593495934, | |
| "grad_norm": 14.820103645324707, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 2.255, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5067750677506775, | |
| "grad_norm": 9.711482048034668, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 3.0789, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5176151761517616, | |
| "grad_norm": 10.218128204345703, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 3.092, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5284552845528454, | |
| "grad_norm": 11.111710548400879, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 3.1396, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.5392953929539295, | |
| "grad_norm": 13.117948532104492, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 2.7166, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.5501355013550135, | |
| "grad_norm": 18.812355041503906, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.6296, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.5609756097560976, | |
| "grad_norm": 11.009380340576172, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 2.5875, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.5718157181571817, | |
| "grad_norm": 12.999532699584961, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 2.739, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.5826558265582655, | |
| "grad_norm": 11.887777328491211, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 2.5411, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.5934959349593496, | |
| "grad_norm": 11.036377906799316, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 2.1869, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6043360433604335, | |
| "grad_norm": 12.381731033325195, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 2.409, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6151761517615175, | |
| "grad_norm": 10.615528106689453, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 2.35, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6260162601626016, | |
| "grad_norm": 10.526694297790527, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 2.3737, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6260162601626016, | |
| "eval_loss": 1.1211493015289307, | |
| "eval_runtime": 11.6468, | |
| "eval_samples_per_second": 13.394, | |
| "eval_steps_per_second": 3.349, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6368563685636857, | |
| "grad_norm": 10.325928688049316, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 2.1218, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.6476964769647697, | |
| "grad_norm": 11.65972900390625, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 2.029, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.6585365853658538, | |
| "grad_norm": 11.227371215820312, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 2.1346, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.6693766937669376, | |
| "grad_norm": 10.244525909423828, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 2.3587, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.6802168021680217, | |
| "grad_norm": 11.884202003479004, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 2.543, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.6910569105691056, | |
| "grad_norm": 10.535400390625, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 2.1798, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7018970189701896, | |
| "grad_norm": 11.714420318603516, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 2.401, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7127371273712737, | |
| "grad_norm": 10.614241600036621, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 1.9869, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7235772357723578, | |
| "grad_norm": 12.482707023620605, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 2.0333, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.7344173441734418, | |
| "grad_norm": 10.333512306213379, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 1.5769, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.7452574525745257, | |
| "grad_norm": 12.281250953674316, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 1.8421, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.7560975609756098, | |
| "grad_norm": 9.520055770874023, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 3.6283, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.7669376693766936, | |
| "grad_norm": 9.745546340942383, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 3.1176, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 12.058420181274414, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 2.6092, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.7886178861788617, | |
| "grad_norm": 11.818090438842773, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 2.8085, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.7994579945799458, | |
| "grad_norm": 10.319408416748047, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 2.7864, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8102981029810299, | |
| "grad_norm": 9.980995178222656, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 2.4463, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.821138211382114, | |
| "grad_norm": 11.36983871459961, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 2.4133, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.8319783197831978, | |
| "grad_norm": 10.295164108276367, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 2.4036, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.8428184281842819, | |
| "grad_norm": 10.947744369506836, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 2.4432, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.8536585365853657, | |
| "grad_norm": 10.340069770812988, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 2.0875, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.8644986449864498, | |
| "grad_norm": 8.746782302856445, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 1.9539, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.8753387533875339, | |
| "grad_norm": 10.641772270202637, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 2.398, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.886178861788618, | |
| "grad_norm": 9.910839080810547, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 2.0354, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.897018970189702, | |
| "grad_norm": 11.399153709411621, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 2.2897, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.907859078590786, | |
| "grad_norm": 11.40278148651123, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 2.0488, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.91869918699187, | |
| "grad_norm": 10.247564315795898, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 2.0313, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.9295392953929538, | |
| "grad_norm": 11.34393310546875, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 2.0246, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.9403794037940378, | |
| "grad_norm": 12.154601097106934, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 2.229, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.951219512195122, | |
| "grad_norm": 10.95048713684082, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 2.1193, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.962059620596206, | |
| "grad_norm": 11.034046173095703, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.9931, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.97289972899729, | |
| "grad_norm": 13.242072105407715, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 2.2749, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.9837398373983741, | |
| "grad_norm": 12.907384872436523, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 1.9783, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.994579945799458, | |
| "grad_norm": 12.940130233764648, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 1.9405, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.005420054200542, | |
| "grad_norm": 16.93910789489746, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 2.896, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.016260162601626, | |
| "grad_norm": 6.427479267120361, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 1.9814, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.02710027100271, | |
| "grad_norm": 7.64243221282959, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 1.8503, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.037940379403794, | |
| "grad_norm": 6.146543979644775, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 1.8137, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.048780487804878, | |
| "grad_norm": 6.6822896003723145, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 1.489, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.059620596205962, | |
| "grad_norm": 7.969797611236572, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 1.728, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.0704607046070462, | |
| "grad_norm": 7.673564910888672, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 1.6888, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.08130081300813, | |
| "grad_norm": 7.841696262359619, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 1.5244, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.092140921409214, | |
| "grad_norm": 7.968762397766113, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 1.6366, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.102981029810298, | |
| "grad_norm": 6.927262783050537, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 1.3177, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.113821138211382, | |
| "grad_norm": 7.036567687988281, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 1.497, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.124661246612466, | |
| "grad_norm": 14.829819679260254, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 1.4145, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.13550135501355, | |
| "grad_norm": 7.1791181564331055, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 1.2565, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.1463414634146343, | |
| "grad_norm": 7.746449947357178, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 1.2621, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.1571815718157183, | |
| "grad_norm": 7.811188697814941, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 1.4829, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.168021680216802, | |
| "grad_norm": 6.5562615394592285, | |
| "learning_rate": 0.0, | |
| "loss": 1.2563, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.168021680216802, | |
| "eval_loss": 1.116738200187683, | |
| "eval_runtime": 11.651, | |
| "eval_samples_per_second": 13.389, | |
| "eval_steps_per_second": 3.347, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "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": 2.8557386158650163e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |