| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.955414012738854, | |
| "eval_steps": 500, | |
| "global_step": 364, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01910828025477707, | |
| "grad_norm": 15.085136704148589, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 1.5704, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03821656050955414, | |
| "grad_norm": 14.474938251005387, | |
| "learning_rate": 1.0810810810810812e-06, | |
| "loss": 1.5563, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05732484076433121, | |
| "grad_norm": 15.32568517104267, | |
| "learning_rate": 1.6216216216216219e-06, | |
| "loss": 1.5849, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07643312101910828, | |
| "grad_norm": 14.428571712044095, | |
| "learning_rate": 2.1621621621621623e-06, | |
| "loss": 1.5303, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09554140127388536, | |
| "grad_norm": 13.896963510791291, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 1.5134, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11464968152866242, | |
| "grad_norm": 11.006953072759748, | |
| "learning_rate": 3.2432432432432437e-06, | |
| "loss": 1.4081, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1337579617834395, | |
| "grad_norm": 9.206149302415982, | |
| "learning_rate": 3.7837837837837844e-06, | |
| "loss": 1.2991, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.15286624203821655, | |
| "grad_norm": 6.94677073601279, | |
| "learning_rate": 4.324324324324325e-06, | |
| "loss": 1.2721, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.17197452229299362, | |
| "grad_norm": 5.901205531632909, | |
| "learning_rate": 4.864864864864866e-06, | |
| "loss": 1.2642, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1910828025477707, | |
| "grad_norm": 7.287831438076889, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 1.1762, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.21019108280254778, | |
| "grad_norm": 5.620841345608181, | |
| "learning_rate": 5.945945945945947e-06, | |
| "loss": 1.1208, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.22929936305732485, | |
| "grad_norm": 4.642852067366921, | |
| "learning_rate": 6.486486486486487e-06, | |
| "loss": 1.0816, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.2484076433121019, | |
| "grad_norm": 4.3318726246843475, | |
| "learning_rate": 7.027027027027028e-06, | |
| "loss": 1.0036, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.267515923566879, | |
| "grad_norm": 5.0994629088458545, | |
| "learning_rate": 7.567567567567569e-06, | |
| "loss": 1.0178, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.28662420382165604, | |
| "grad_norm": 4.7221655440052706, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 1.0188, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3057324840764331, | |
| "grad_norm": 4.035891732049943, | |
| "learning_rate": 8.64864864864865e-06, | |
| "loss": 0.9837, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3248407643312102, | |
| "grad_norm": 2.340655624907049, | |
| "learning_rate": 9.189189189189191e-06, | |
| "loss": 0.9815, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.34394904458598724, | |
| "grad_norm": 2.937006665479249, | |
| "learning_rate": 9.729729729729732e-06, | |
| "loss": 0.9641, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3630573248407643, | |
| "grad_norm": 3.256335346482645, | |
| "learning_rate": 1.027027027027027e-05, | |
| "loss": 0.945, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3821656050955414, | |
| "grad_norm": 3.3901721436513514, | |
| "learning_rate": 1.0810810810810812e-05, | |
| "loss": 0.9176, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4012738853503185, | |
| "grad_norm": 2.830848553381318, | |
| "learning_rate": 1.1351351351351352e-05, | |
| "loss": 0.9487, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.42038216560509556, | |
| "grad_norm": 2.3955522880864306, | |
| "learning_rate": 1.1891891891891894e-05, | |
| "loss": 0.9298, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.4394904458598726, | |
| "grad_norm": 2.3058955393493785, | |
| "learning_rate": 1.2432432432432433e-05, | |
| "loss": 0.9028, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4585987261146497, | |
| "grad_norm": 2.2249716284688077, | |
| "learning_rate": 1.2972972972972975e-05, | |
| "loss": 0.9011, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 1.8607602104830938, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 0.8792, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4968152866242038, | |
| "grad_norm": 1.8628091456429328, | |
| "learning_rate": 1.4054054054054055e-05, | |
| "loss": 0.8868, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5159235668789809, | |
| "grad_norm": 1.872436926273032, | |
| "learning_rate": 1.4594594594594596e-05, | |
| "loss": 0.8601, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.535031847133758, | |
| "grad_norm": 1.8178491230546292, | |
| "learning_rate": 1.5135135135135138e-05, | |
| "loss": 0.8504, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.554140127388535, | |
| "grad_norm": 2.0301685866957446, | |
| "learning_rate": 1.5675675675675676e-05, | |
| "loss": 0.8594, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5732484076433121, | |
| "grad_norm": 1.6561209868585802, | |
| "learning_rate": 1.6216216216216218e-05, | |
| "loss": 0.8806, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5923566878980892, | |
| "grad_norm": 1.8719033101162887, | |
| "learning_rate": 1.6756756756756757e-05, | |
| "loss": 0.8396, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6114649681528662, | |
| "grad_norm": 1.8213542449373956, | |
| "learning_rate": 1.72972972972973e-05, | |
| "loss": 0.8556, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.6305732484076433, | |
| "grad_norm": 2.155963366674303, | |
| "learning_rate": 1.783783783783784e-05, | |
| "loss": 0.8606, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6496815286624203, | |
| "grad_norm": 1.7402657268229764, | |
| "learning_rate": 1.8378378378378383e-05, | |
| "loss": 0.8676, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6687898089171974, | |
| "grad_norm": 1.8170627228487601, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 0.8417, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6878980891719745, | |
| "grad_norm": 1.7896656829907667, | |
| "learning_rate": 1.9459459459459463e-05, | |
| "loss": 0.8403, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7070063694267515, | |
| "grad_norm": 1.9598160171117724, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8542, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7261146496815286, | |
| "grad_norm": 1.805888927732859, | |
| "learning_rate": 1.9999538500851633e-05, | |
| "loss": 0.8497, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7452229299363057, | |
| "grad_norm": 1.7523676716761967, | |
| "learning_rate": 1.9998154046002822e-05, | |
| "loss": 0.8226, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7643312101910829, | |
| "grad_norm": 1.8417462488969998, | |
| "learning_rate": 1.9995846763238514e-05, | |
| "loss": 0.8387, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7834394904458599, | |
| "grad_norm": 2.1434813652680234, | |
| "learning_rate": 1.9992616865520515e-05, | |
| "loss": 0.8221, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.802547770700637, | |
| "grad_norm": 1.6245041662270168, | |
| "learning_rate": 1.9988464650967834e-05, | |
| "loss": 0.829, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.821656050955414, | |
| "grad_norm": 1.7984749274517189, | |
| "learning_rate": 1.9983390502829168e-05, | |
| "loss": 0.832, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8407643312101911, | |
| "grad_norm": 1.9112707656400993, | |
| "learning_rate": 1.9977394889447526e-05, | |
| "loss": 0.8285, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8598726114649682, | |
| "grad_norm": 1.7163027516640725, | |
| "learning_rate": 1.9970478364217e-05, | |
| "loss": 0.8239, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8789808917197452, | |
| "grad_norm": 1.640177988055144, | |
| "learning_rate": 1.9962641565531694e-05, | |
| "loss": 0.839, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8980891719745223, | |
| "grad_norm": 1.6204664057576417, | |
| "learning_rate": 1.9953885216726788e-05, | |
| "loss": 0.8366, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.9171974522292994, | |
| "grad_norm": 1.6305143686553394, | |
| "learning_rate": 1.994421012601179e-05, | |
| "loss": 0.8033, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.9363057324840764, | |
| "grad_norm": 1.645454323245968, | |
| "learning_rate": 1.9933617186395917e-05, | |
| "loss": 0.7967, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 1.4427057124607237, | |
| "learning_rate": 1.99221073756057e-05, | |
| "loss": 0.824, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9745222929936306, | |
| "grad_norm": 1.6145093346377806, | |
| "learning_rate": 1.990968175599471e-05, | |
| "loss": 0.8158, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9936305732484076, | |
| "grad_norm": 1.5999720545933496, | |
| "learning_rate": 1.9896341474445526e-05, | |
| "loss": 0.8219, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.0127388535031847, | |
| "grad_norm": 1.5556656213210873, | |
| "learning_rate": 1.9882087762263857e-05, | |
| "loss": 0.6919, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.0318471337579618, | |
| "grad_norm": 1.2836749011363615, | |
| "learning_rate": 1.9866921935064907e-05, | |
| "loss": 0.6448, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0509554140127388, | |
| "grad_norm": 1.3220942267631135, | |
| "learning_rate": 1.985084539265195e-05, | |
| "loss": 0.6314, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.070063694267516, | |
| "grad_norm": 1.602108008943959, | |
| "learning_rate": 1.983385961888711e-05, | |
| "loss": 0.6361, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.089171974522293, | |
| "grad_norm": 1.394790824644809, | |
| "learning_rate": 1.9815966181554412e-05, | |
| "loss": 0.6109, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.10828025477707, | |
| "grad_norm": 1.6712253019175645, | |
| "learning_rate": 1.9797166732215078e-05, | |
| "loss": 0.6451, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.127388535031847, | |
| "grad_norm": 1.5616106813163366, | |
| "learning_rate": 1.977746300605507e-05, | |
| "loss": 0.6296, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.1464968152866242, | |
| "grad_norm": 1.6518390030361483, | |
| "learning_rate": 1.975685682172497e-05, | |
| "loss": 0.6108, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1656050955414012, | |
| "grad_norm": 1.5709075226406595, | |
| "learning_rate": 1.973535008117207e-05, | |
| "loss": 0.6578, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.1847133757961783, | |
| "grad_norm": 1.558384294002867, | |
| "learning_rate": 1.9712944769464864e-05, | |
| "loss": 0.6217, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.2038216560509554, | |
| "grad_norm": 1.4728288614515401, | |
| "learning_rate": 1.9689642954609808e-05, | |
| "loss": 0.6282, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.2229299363057324, | |
| "grad_norm": 1.4521431261528124, | |
| "learning_rate": 1.9665446787360444e-05, | |
| "loss": 0.6222, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.2420382165605095, | |
| "grad_norm": 1.349921330334354, | |
| "learning_rate": 1.9640358501018885e-05, | |
| "loss": 0.6181, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.2611464968152866, | |
| "grad_norm": 1.6237778029281429, | |
| "learning_rate": 1.9614380411229693e-05, | |
| "loss": 0.6513, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2802547770700636, | |
| "grad_norm": 1.2105413945238757, | |
| "learning_rate": 1.9587514915766124e-05, | |
| "loss": 0.6044, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.2993630573248407, | |
| "grad_norm": 1.550519429214243, | |
| "learning_rate": 1.9559764494308838e-05, | |
| "loss": 0.6014, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.3184713375796178, | |
| "grad_norm": 1.3407009038102518, | |
| "learning_rate": 1.9531131708217005e-05, | |
| "loss": 0.6547, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.3375796178343948, | |
| "grad_norm": 1.4103946434179544, | |
| "learning_rate": 1.950161920029191e-05, | |
| "loss": 0.582, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.356687898089172, | |
| "grad_norm": 1.4628448230037054, | |
| "learning_rate": 1.9471229694533003e-05, | |
| "loss": 0.5937, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.3757961783439492, | |
| "grad_norm": 1.5348702170741033, | |
| "learning_rate": 1.943996599588649e-05, | |
| "loss": 0.6058, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.394904458598726, | |
| "grad_norm": 1.4250118657179596, | |
| "learning_rate": 1.940783098998643e-05, | |
| "loss": 0.6032, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.4140127388535033, | |
| "grad_norm": 1.7735403263967013, | |
| "learning_rate": 1.93748276428884e-05, | |
| "loss": 0.6206, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 1.4297378037077724, | |
| "learning_rate": 1.9340959000795707e-05, | |
| "loss": 0.5988, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4522292993630574, | |
| "grad_norm": 1.5220436230214494, | |
| "learning_rate": 1.9306228189778255e-05, | |
| "loss": 0.6123, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.4713375796178343, | |
| "grad_norm": 1.616878937280132, | |
| "learning_rate": 1.927063841548398e-05, | |
| "loss": 0.5997, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.4904458598726116, | |
| "grad_norm": 1.34310202463036, | |
| "learning_rate": 1.9234192962842996e-05, | |
| "loss": 0.619, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.5095541401273884, | |
| "grad_norm": 1.4229036274818758, | |
| "learning_rate": 1.9196895195764363e-05, | |
| "loss": 0.6017, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.5286624203821657, | |
| "grad_norm": 1.390423276264005, | |
| "learning_rate": 1.9158748556825637e-05, | |
| "loss": 0.5528, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.5477707006369426, | |
| "grad_norm": 1.258187913754368, | |
| "learning_rate": 1.9119756566955092e-05, | |
| "loss": 0.5987, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.5668789808917198, | |
| "grad_norm": 1.4948917254001746, | |
| "learning_rate": 1.907992282510675e-05, | |
| "loss": 0.6284, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.5859872611464967, | |
| "grad_norm": 1.4424358931456185, | |
| "learning_rate": 1.90392510079282e-05, | |
| "loss": 0.6186, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.605095541401274, | |
| "grad_norm": 1.2523402042193517, | |
| "learning_rate": 1.8997744869421248e-05, | |
| "loss": 0.6302, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.6242038216560508, | |
| "grad_norm": 1.3303361792693773, | |
| "learning_rate": 1.8955408240595396e-05, | |
| "loss": 0.6261, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.643312101910828, | |
| "grad_norm": 1.4494251986517679, | |
| "learning_rate": 1.891224502911428e-05, | |
| "loss": 0.6019, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.662420382165605, | |
| "grad_norm": 1.4290430974733064, | |
| "learning_rate": 1.886825921893497e-05, | |
| "loss": 0.6023, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.6815286624203822, | |
| "grad_norm": 1.4144543041646682, | |
| "learning_rate": 1.8823454869940243e-05, | |
| "loss": 0.6175, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.700636942675159, | |
| "grad_norm": 1.288956801783328, | |
| "learning_rate": 1.8777836117563894e-05, | |
| "loss": 0.5923, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.7197452229299364, | |
| "grad_norm": 1.464201550337535, | |
| "learning_rate": 1.873140717240899e-05, | |
| "loss": 0.6155, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.7388535031847132, | |
| "grad_norm": 1.3813985093874788, | |
| "learning_rate": 1.8684172319859258e-05, | |
| "loss": 0.5979, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.7579617834394905, | |
| "grad_norm": 1.2688637001507534, | |
| "learning_rate": 1.863613591968355e-05, | |
| "loss": 0.6105, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.7770700636942676, | |
| "grad_norm": 1.2970848534467883, | |
| "learning_rate": 1.858730240563342e-05, | |
| "loss": 0.6227, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.7961783439490446, | |
| "grad_norm": 1.448667194735226, | |
| "learning_rate": 1.8537676285033886e-05, | |
| "loss": 0.6112, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.8152866242038217, | |
| "grad_norm": 1.306025010226125, | |
| "learning_rate": 1.848726213836744e-05, | |
| "loss": 0.6264, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.8343949044585988, | |
| "grad_norm": 1.3231748537558505, | |
| "learning_rate": 1.8436064618851225e-05, | |
| "loss": 0.6083, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.8535031847133758, | |
| "grad_norm": 1.4183975917076637, | |
| "learning_rate": 1.838408845200758e-05, | |
| "loss": 0.617, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.872611464968153, | |
| "grad_norm": 1.2221404350574396, | |
| "learning_rate": 1.8331338435227838e-05, | |
| "loss": 0.6111, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.89171974522293, | |
| "grad_norm": 1.2866603816546331, | |
| "learning_rate": 1.8277819437329577e-05, | |
| "loss": 0.6087, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 1.4005424490311182, | |
| "learning_rate": 1.8223536398107177e-05, | |
| "loss": 0.6211, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.929936305732484, | |
| "grad_norm": 1.3404707899448898, | |
| "learning_rate": 1.8168494327875918e-05, | |
| "loss": 0.6467, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.9490445859872612, | |
| "grad_norm": 1.3904382526949657, | |
| "learning_rate": 1.8112698307009506e-05, | |
| "loss": 0.6247, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.9681528662420382, | |
| "grad_norm": 1.3475220635894605, | |
| "learning_rate": 1.8056153485471167e-05, | |
| "loss": 0.5951, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.9872611464968153, | |
| "grad_norm": 1.2503632231175987, | |
| "learning_rate": 1.799886508233829e-05, | |
| "loss": 0.6204, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.0063694267515926, | |
| "grad_norm": 1.5062759250906272, | |
| "learning_rate": 1.7940838385320732e-05, | |
| "loss": 0.4785, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.0254777070063694, | |
| "grad_norm": 1.2957519949624863, | |
| "learning_rate": 1.788207875027274e-05, | |
| "loss": 0.3498, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.0445859872611467, | |
| "grad_norm": 1.7472953059783831, | |
| "learning_rate": 1.7822591600698632e-05, | |
| "loss": 0.3749, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.0636942675159236, | |
| "grad_norm": 1.8419893598127797, | |
| "learning_rate": 1.776238242725217e-05, | |
| "loss": 0.381, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.082802547770701, | |
| "grad_norm": 1.2418920063978576, | |
| "learning_rate": 1.7701456787229805e-05, | |
| "loss": 0.3652, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.1019108280254777, | |
| "grad_norm": 1.2669456195137885, | |
| "learning_rate": 1.7639820304057745e-05, | |
| "loss": 0.3626, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.121019108280255, | |
| "grad_norm": 1.3210572020344045, | |
| "learning_rate": 1.7577478666772886e-05, | |
| "loss": 0.3552, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.140127388535032, | |
| "grad_norm": 1.2878579307569704, | |
| "learning_rate": 1.751443762949772e-05, | |
| "loss": 0.3916, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.159235668789809, | |
| "grad_norm": 1.0835624768946444, | |
| "learning_rate": 1.7450703010909263e-05, | |
| "loss": 0.3522, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.178343949044586, | |
| "grad_norm": 1.3141122312855134, | |
| "learning_rate": 1.738628069370195e-05, | |
| "loss": 0.3564, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.1974522292993632, | |
| "grad_norm": 1.1884378897572276, | |
| "learning_rate": 1.732117662404469e-05, | |
| "loss": 0.3824, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.21656050955414, | |
| "grad_norm": 1.2916282158570314, | |
| "learning_rate": 1.7255396811032014e-05, | |
| "loss": 0.3361, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.2356687898089174, | |
| "grad_norm": 1.16206821276476, | |
| "learning_rate": 1.718894732612947e-05, | |
| "loss": 0.3992, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.254777070063694, | |
| "grad_norm": 1.2002118548011045, | |
| "learning_rate": 1.712183430261319e-05, | |
| "loss": 0.3331, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.2738853503184715, | |
| "grad_norm": 1.238038592697249, | |
| "learning_rate": 1.7054063935003813e-05, | |
| "loss": 0.3637, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.2929936305732483, | |
| "grad_norm": 1.276468965169396, | |
| "learning_rate": 1.698564247849473e-05, | |
| "loss": 0.3561, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.3121019108280256, | |
| "grad_norm": 1.1858057882929658, | |
| "learning_rate": 1.691657624837472e-05, | |
| "loss": 0.3562, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.3312101910828025, | |
| "grad_norm": 1.1772039337987994, | |
| "learning_rate": 1.684687161944506e-05, | |
| "loss": 0.3741, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.3503184713375798, | |
| "grad_norm": 1.191796976393949, | |
| "learning_rate": 1.677653502543113e-05, | |
| "loss": 0.3149, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.3694267515923566, | |
| "grad_norm": 1.1834181104795152, | |
| "learning_rate": 1.6705572958388576e-05, | |
| "loss": 0.348, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 1.2225814732311175, | |
| "learning_rate": 1.6633991968104095e-05, | |
| "loss": 0.327, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.4076433121019107, | |
| "grad_norm": 1.1447586367316194, | |
| "learning_rate": 1.6561798661490904e-05, | |
| "loss": 0.3458, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.426751592356688, | |
| "grad_norm": 1.1335284502819871, | |
| "learning_rate": 1.6488999701978905e-05, | |
| "loss": 0.3672, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.445859872611465, | |
| "grad_norm": 1.2270270565838597, | |
| "learning_rate": 1.6415601808899658e-05, | |
| "loss": 0.3518, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.464968152866242, | |
| "grad_norm": 1.1899082443586815, | |
| "learning_rate": 1.63416117568662e-05, | |
| "loss": 0.3732, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.484076433121019, | |
| "grad_norm": 1.1501300313912142, | |
| "learning_rate": 1.6267036375147728e-05, | |
| "loss": 0.3385, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.5031847133757963, | |
| "grad_norm": 1.1375198551413968, | |
| "learning_rate": 1.619188254703927e-05, | |
| "loss": 0.3489, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.522292993630573, | |
| "grad_norm": 1.1458583825980333, | |
| "learning_rate": 1.6116157209226356e-05, | |
| "loss": 0.3449, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.5414012738853504, | |
| "grad_norm": 1.1186369151785944, | |
| "learning_rate": 1.6039867351144778e-05, | |
| "loss": 0.3771, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.5605095541401273, | |
| "grad_norm": 1.0755679290761608, | |
| "learning_rate": 1.5963020014335437e-05, | |
| "loss": 0.3599, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.5796178343949046, | |
| "grad_norm": 1.0442201841536443, | |
| "learning_rate": 1.588562229179443e-05, | |
| "loss": 0.3738, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.5987261146496814, | |
| "grad_norm": 1.0954097602146617, | |
| "learning_rate": 1.5807681327318372e-05, | |
| "loss": 0.3118, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.6178343949044587, | |
| "grad_norm": 1.1765113644935257, | |
| "learning_rate": 1.5729204314845002e-05, | |
| "loss": 0.3226, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.6369426751592355, | |
| "grad_norm": 1.175408508344713, | |
| "learning_rate": 1.56501984977892e-05, | |
| "loss": 0.3595, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.656050955414013, | |
| "grad_norm": 1.145496103535334, | |
| "learning_rate": 1.557067116837444e-05, | |
| "loss": 0.3403, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.6751592356687897, | |
| "grad_norm": 1.0853612573171887, | |
| "learning_rate": 1.5490629666959668e-05, | |
| "loss": 0.3686, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.694267515923567, | |
| "grad_norm": 1.1754047814510593, | |
| "learning_rate": 1.541008138136183e-05, | |
| "loss": 0.3658, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.713375796178344, | |
| "grad_norm": 1.2715689153818588, | |
| "learning_rate": 1.5329033746173975e-05, | |
| "loss": 0.3784, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.732484076433121, | |
| "grad_norm": 1.1204039055112898, | |
| "learning_rate": 1.5247494242079024e-05, | |
| "loss": 0.399, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.7515923566878984, | |
| "grad_norm": 1.172167778228388, | |
| "learning_rate": 1.5165470395159314e-05, | |
| "loss": 0.3555, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.770700636942675, | |
| "grad_norm": 1.1808325543649851, | |
| "learning_rate": 1.5082969776201948e-05, | |
| "loss": 0.3641, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.789808917197452, | |
| "grad_norm": 1.2556255666144898, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3908, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.8089171974522293, | |
| "grad_norm": 1.0250278920355878, | |
| "learning_rate": 1.4916568724649688e-05, | |
| "loss": 0.3919, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.8280254777070066, | |
| "grad_norm": 1.0885042466878538, | |
| "learning_rate": 1.483268365084351e-05, | |
| "loss": 0.3675, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.8471337579617835, | |
| "grad_norm": 1.0769198665896773, | |
| "learning_rate": 1.4748352521159492e-05, | |
| "loss": 0.3366, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 1.0867086822203522, | |
| "learning_rate": 1.466358311934654e-05, | |
| "loss": 0.3745, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.8853503184713376, | |
| "grad_norm": 1.1682869149729933, | |
| "learning_rate": 1.4578383269606004e-05, | |
| "loss": 0.3584, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.904458598726115, | |
| "grad_norm": 1.0638669162017096, | |
| "learning_rate": 1.4492760835869504e-05, | |
| "loss": 0.3649, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.9235668789808917, | |
| "grad_norm": 1.207017922975427, | |
| "learning_rate": 1.4406723721073088e-05, | |
| "loss": 0.3988, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.9426751592356686, | |
| "grad_norm": 1.0877999612365532, | |
| "learning_rate": 1.4320279866427798e-05, | |
| "loss": 0.3687, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.961783439490446, | |
| "grad_norm": 1.0387216426253332, | |
| "learning_rate": 1.4233437250686695e-05, | |
| "loss": 0.3845, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.980891719745223, | |
| "grad_norm": 1.0977652974730174, | |
| "learning_rate": 1.4146203889408418e-05, | |
| "loss": 0.3706, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0055100599636702, | |
| "learning_rate": 1.4058587834217356e-05, | |
| "loss": 0.3452, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 3.0191082802547773, | |
| "grad_norm": 1.0618160168070347, | |
| "learning_rate": 1.3970597172060482e-05, | |
| "loss": 0.2098, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.038216560509554, | |
| "grad_norm": 0.8998396967279327, | |
| "learning_rate": 1.3882240024460928e-05, | |
| "loss": 0.2096, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 3.0573248407643314, | |
| "grad_norm": 0.9702115715621121, | |
| "learning_rate": 1.3793524546768358e-05, | |
| "loss": 0.2091, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.0764331210191083, | |
| "grad_norm": 1.308777911334755, | |
| "learning_rate": 1.3704458927406261e-05, | |
| "loss": 0.2332, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 3.0955414012738856, | |
| "grad_norm": 1.17659608000573, | |
| "learning_rate": 1.3615051387116131e-05, | |
| "loss": 0.2284, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 3.1146496815286624, | |
| "grad_norm": 1.011176682188353, | |
| "learning_rate": 1.3525310178198707e-05, | |
| "loss": 0.24, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 3.1337579617834397, | |
| "grad_norm": 0.9604835740358206, | |
| "learning_rate": 1.3435243583752294e-05, | |
| "loss": 0.1868, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 3.1528662420382165, | |
| "grad_norm": 1.028174294533024, | |
| "learning_rate": 1.3344859916908206e-05, | |
| "loss": 0.2063, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.171974522292994, | |
| "grad_norm": 1.0037868388480298, | |
| "learning_rate": 1.325416752006351e-05, | |
| "loss": 0.2199, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.1910828025477707, | |
| "grad_norm": 0.873374998586254, | |
| "learning_rate": 1.3163174764110985e-05, | |
| "loss": 0.2307, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 3.210191082802548, | |
| "grad_norm": 0.9301743491679733, | |
| "learning_rate": 1.3071890047666498e-05, | |
| "loss": 0.1969, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 3.229299363057325, | |
| "grad_norm": 1.0223903303715116, | |
| "learning_rate": 1.2980321796293838e-05, | |
| "loss": 0.2196, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 3.248407643312102, | |
| "grad_norm": 0.9204748338830222, | |
| "learning_rate": 1.288847846172701e-05, | |
| "loss": 0.209, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.267515923566879, | |
| "grad_norm": 0.9700392345403084, | |
| "learning_rate": 1.2796368521090143e-05, | |
| "loss": 0.2395, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.286624203821656, | |
| "grad_norm": 0.8599148705622103, | |
| "learning_rate": 1.2704000476115079e-05, | |
| "loss": 0.2369, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.305732484076433, | |
| "grad_norm": 0.9114872515780617, | |
| "learning_rate": 1.2611382852356632e-05, | |
| "loss": 0.2145, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.3248407643312103, | |
| "grad_norm": 0.9430415582365258, | |
| "learning_rate": 1.2518524198405699e-05, | |
| "loss": 0.2099, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.343949044585987, | |
| "grad_norm": 0.9000866161280985, | |
| "learning_rate": 1.2425433085100224e-05, | |
| "loss": 0.1996, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.3630573248407645, | |
| "grad_norm": 0.8677434563082431, | |
| "learning_rate": 1.233211810473411e-05, | |
| "loss": 0.2061, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.3821656050955413, | |
| "grad_norm": 0.9008387292004424, | |
| "learning_rate": 1.2238587870264152e-05, | |
| "loss": 0.2279, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.4012738853503186, | |
| "grad_norm": 0.8778321737636415, | |
| "learning_rate": 1.2144851014515055e-05, | |
| "loss": 0.2063, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.4203821656050954, | |
| "grad_norm": 0.8948567421126848, | |
| "learning_rate": 1.2050916189382646e-05, | |
| "loss": 0.2352, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.4394904458598727, | |
| "grad_norm": 0.8664926126254838, | |
| "learning_rate": 1.1956792065035281e-05, | |
| "loss": 0.2176, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.4585987261146496, | |
| "grad_norm": 0.8812689097943135, | |
| "learning_rate": 1.1862487329113606e-05, | |
| "loss": 0.2075, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.477707006369427, | |
| "grad_norm": 0.8417966696612939, | |
| "learning_rate": 1.1768010685928686e-05, | |
| "loss": 0.2328, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.4968152866242037, | |
| "grad_norm": 0.9307534883666011, | |
| "learning_rate": 1.1673370855658592e-05, | |
| "loss": 0.2345, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.515923566878981, | |
| "grad_norm": 0.9310328230598713, | |
| "learning_rate": 1.1578576573543541e-05, | |
| "loss": 0.1951, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.535031847133758, | |
| "grad_norm": 0.8610138134793425, | |
| "learning_rate": 1.1483636589079627e-05, | |
| "loss": 0.2279, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.554140127388535, | |
| "grad_norm": 0.9547154541149019, | |
| "learning_rate": 1.1388559665211241e-05, | |
| "loss": 0.2219, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.573248407643312, | |
| "grad_norm": 0.8593407029066408, | |
| "learning_rate": 1.1293354577522264e-05, | |
| "loss": 0.2308, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 3.5923566878980893, | |
| "grad_norm": 0.8967627254818533, | |
| "learning_rate": 1.1198030113426074e-05, | |
| "loss": 0.2068, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.611464968152866, | |
| "grad_norm": 0.8982632517352984, | |
| "learning_rate": 1.1102595071354471e-05, | |
| "loss": 0.2206, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 3.6305732484076434, | |
| "grad_norm": 1.011950144597425, | |
| "learning_rate": 1.1007058259945584e-05, | |
| "loss": 0.2271, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.6496815286624202, | |
| "grad_norm": 0.9110300503235955, | |
| "learning_rate": 1.0911428497230834e-05, | |
| "loss": 0.1933, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.6687898089171975, | |
| "grad_norm": 0.8902196629233927, | |
| "learning_rate": 1.0815714609821027e-05, | |
| "loss": 0.1926, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.6878980891719744, | |
| "grad_norm": 0.8903961297865141, | |
| "learning_rate": 1.0719925432091671e-05, | |
| "loss": 0.2226, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.7070063694267517, | |
| "grad_norm": 0.8911238784979923, | |
| "learning_rate": 1.0624069805367558e-05, | |
| "loss": 0.2285, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.7261146496815285, | |
| "grad_norm": 0.9483518994037937, | |
| "learning_rate": 1.0528156577106703e-05, | |
| "loss": 0.2378, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.745222929936306, | |
| "grad_norm": 0.9136627627640508, | |
| "learning_rate": 1.043219460008374e-05, | |
| "loss": 0.2319, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.7643312101910826, | |
| "grad_norm": 0.8894840057415601, | |
| "learning_rate": 1.0336192731572805e-05, | |
| "loss": 0.1874, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.78343949044586, | |
| "grad_norm": 0.902747172951627, | |
| "learning_rate": 1.0240159832530007e-05, | |
| "loss": 0.2154, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.802547770700637, | |
| "grad_norm": 0.8890841549619691, | |
| "learning_rate": 1.0144104766775574e-05, | |
| "loss": 0.2287, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 0.8341873329824605, | |
| "learning_rate": 1.004803640017571e-05, | |
| "loss": 0.2018, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.840764331210191, | |
| "grad_norm": 0.8978791085460874, | |
| "learning_rate": 9.951963599824294e-06, | |
| "loss": 0.2283, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.859872611464968, | |
| "grad_norm": 0.8971165967335231, | |
| "learning_rate": 9.855895233224431e-06, | |
| "loss": 0.2092, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 3.8789808917197455, | |
| "grad_norm": 0.884773612492414, | |
| "learning_rate": 9.759840167469995e-06, | |
| "loss": 0.1998, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.8980891719745223, | |
| "grad_norm": 0.881331674767945, | |
| "learning_rate": 9.663807268427197e-06, | |
| "loss": 0.2107, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.917197452229299, | |
| "grad_norm": 0.8694706950976717, | |
| "learning_rate": 9.56780539991626e-06, | |
| "loss": 0.1737, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.9363057324840764, | |
| "grad_norm": 0.8745893448622579, | |
| "learning_rate": 9.471843422893299e-06, | |
| "loss": 0.2417, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 3.9554140127388537, | |
| "grad_norm": 0.8547548875079425, | |
| "learning_rate": 9.375930194632447e-06, | |
| "loss": 0.191, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 3.9745222929936306, | |
| "grad_norm": 0.8425865257249392, | |
| "learning_rate": 9.28007456790833e-06, | |
| "loss": 0.2133, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 3.9936305732484074, | |
| "grad_norm": 0.8399199602073947, | |
| "learning_rate": 9.184285390178978e-06, | |
| "loss": 0.1727, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 4.012738853503185, | |
| "grad_norm": 0.6955669225356614, | |
| "learning_rate": 9.08857150276917e-06, | |
| "loss": 0.1507, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.031847133757962, | |
| "grad_norm": 0.7500722614399656, | |
| "learning_rate": 8.992941740054418e-06, | |
| "loss": 0.1424, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 4.050955414012739, | |
| "grad_norm": 0.6750285192172861, | |
| "learning_rate": 8.897404928645529e-06, | |
| "loss": 0.1385, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 4.070063694267516, | |
| "grad_norm": 0.6488864225533145, | |
| "learning_rate": 8.80196988657393e-06, | |
| "loss": 0.1208, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 4.089171974522293, | |
| "grad_norm": 0.7747156519005379, | |
| "learning_rate": 8.706645422477739e-06, | |
| "loss": 0.109, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 4.10828025477707, | |
| "grad_norm": 0.8030446812496158, | |
| "learning_rate": 8.611440334788762e-06, | |
| "loss": 0.11, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.127388535031847, | |
| "grad_norm": 0.842974481181232, | |
| "learning_rate": 8.516363410920376e-06, | |
| "loss": 0.1156, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 4.146496815286624, | |
| "grad_norm": 0.7363547302930578, | |
| "learning_rate": 8.42142342645646e-06, | |
| "loss": 0.1218, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 4.165605095541402, | |
| "grad_norm": 0.7253821835045643, | |
| "learning_rate": 8.326629144341408e-06, | |
| "loss": 0.1305, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 4.1847133757961785, | |
| "grad_norm": 0.6719253785732008, | |
| "learning_rate": 8.231989314071318e-06, | |
| "loss": 0.1394, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 4.203821656050955, | |
| "grad_norm": 0.6902965040425239, | |
| "learning_rate": 8.137512670886397e-06, | |
| "loss": 0.1277, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.222929936305732, | |
| "grad_norm": 0.6588533726729626, | |
| "learning_rate": 8.043207934964722e-06, | |
| "loss": 0.1042, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 4.24203821656051, | |
| "grad_norm": 0.688443320527772, | |
| "learning_rate": 7.949083810617358e-06, | |
| "loss": 0.1395, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 4.261146496815287, | |
| "grad_norm": 0.6848545164428514, | |
| "learning_rate": 7.855148985484946e-06, | |
| "loss": 0.1258, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 4.280254777070064, | |
| "grad_norm": 0.6541473483256922, | |
| "learning_rate": 7.761412129735853e-06, | |
| "loss": 0.1321, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 4.2993630573248405, | |
| "grad_norm": 0.6617548205172367, | |
| "learning_rate": 7.667881895265895e-06, | |
| "loss": 0.1238, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.318471337579618, | |
| "grad_norm": 0.6832552876122532, | |
| "learning_rate": 7.574566914899779e-06, | |
| "loss": 0.1282, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 4.337579617834395, | |
| "grad_norm": 0.7078279469058326, | |
| "learning_rate": 7.481475801594302e-06, | |
| "loss": 0.1498, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 4.356687898089172, | |
| "grad_norm": 0.7054411075095528, | |
| "learning_rate": 7.388617147643371e-06, | |
| "loss": 0.1176, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 4.375796178343949, | |
| "grad_norm": 0.6606922293503963, | |
| "learning_rate": 7.295999523884921e-06, | |
| "loss": 0.1358, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 4.3949044585987265, | |
| "grad_norm": 0.6570134320711147, | |
| "learning_rate": 7.203631478909857e-06, | |
| "loss": 0.1191, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.414012738853503, | |
| "grad_norm": 0.6709587409540739, | |
| "learning_rate": 7.111521538272997e-06, | |
| "loss": 0.1556, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 4.43312101910828, | |
| "grad_norm": 0.6408516242508497, | |
| "learning_rate": 7.019678203706164e-06, | |
| "loss": 0.1145, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 4.452229299363057, | |
| "grad_norm": 0.6310595517744196, | |
| "learning_rate": 6.928109952333506e-06, | |
| "loss": 0.1122, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 4.471337579617835, | |
| "grad_norm": 0.606179060370864, | |
| "learning_rate": 6.83682523588902e-06, | |
| "loss": 0.0894, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 4.490445859872612, | |
| "grad_norm": 0.6482502868898965, | |
| "learning_rate": 6.745832479936492e-06, | |
| "loss": 0.1146, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 4.509554140127388, | |
| "grad_norm": 0.6504239956061556, | |
| "learning_rate": 6.655140083091794e-06, | |
| "loss": 0.1497, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 4.528662420382165, | |
| "grad_norm": 0.6618868734771317, | |
| "learning_rate": 6.564756416247712e-06, | |
| "loss": 0.1388, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 4.547770700636943, | |
| "grad_norm": 0.6575862187025499, | |
| "learning_rate": 6.474689821801295e-06, | |
| "loss": 0.1283, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 4.56687898089172, | |
| "grad_norm": 0.7037809752623807, | |
| "learning_rate": 6.384948612883872e-06, | |
| "loss": 0.1186, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 4.585987261146497, | |
| "grad_norm": 0.6800544654487855, | |
| "learning_rate": 6.2955410725937405e-06, | |
| "loss": 0.134, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.6050955414012735, | |
| "grad_norm": 0.6306670176616782, | |
| "learning_rate": 6.206475453231644e-06, | |
| "loss": 0.1186, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 4.624203821656051, | |
| "grad_norm": 0.6513557094335745, | |
| "learning_rate": 6.117759975539075e-06, | |
| "loss": 0.1181, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 4.643312101910828, | |
| "grad_norm": 0.6277016756429405, | |
| "learning_rate": 6.029402827939519e-06, | |
| "loss": 0.1103, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 4.662420382165605, | |
| "grad_norm": 0.5805626558470603, | |
| "learning_rate": 5.941412165782645e-06, | |
| "loss": 0.1203, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 4.681528662420382, | |
| "grad_norm": 0.6471247978073872, | |
| "learning_rate": 5.853796110591583e-06, | |
| "loss": 0.12, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.7006369426751595, | |
| "grad_norm": 0.6229524298992432, | |
| "learning_rate": 5.766562749313309e-06, | |
| "loss": 0.0914, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 4.719745222929936, | |
| "grad_norm": 0.6171924499516479, | |
| "learning_rate": 5.6797201335722064e-06, | |
| "loss": 0.1129, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 4.738853503184713, | |
| "grad_norm": 0.6360368336026978, | |
| "learning_rate": 5.593276278926912e-06, | |
| "loss": 0.1153, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 4.757961783439491, | |
| "grad_norm": 0.6211962433454916, | |
| "learning_rate": 5.507239164130501e-06, | |
| "loss": 0.1164, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 4.777070063694268, | |
| "grad_norm": 0.629858984427329, | |
| "learning_rate": 5.421616730394e-06, | |
| "loss": 0.1, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.796178343949045, | |
| "grad_norm": 0.6639236503121236, | |
| "learning_rate": 5.336416880653461e-06, | |
| "loss": 0.0996, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 4.8152866242038215, | |
| "grad_norm": 0.6701398769235155, | |
| "learning_rate": 5.251647478840511e-06, | |
| "loss": 0.1347, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.834394904458598, | |
| "grad_norm": 0.6554353086774402, | |
| "learning_rate": 5.167316349156495e-06, | |
| "loss": 0.142, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 4.853503184713376, | |
| "grad_norm": 0.5930407824977478, | |
| "learning_rate": 5.083431275350312e-06, | |
| "loss": 0.1061, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 4.872611464968153, | |
| "grad_norm": 0.6723808750035135, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.1287, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.89171974522293, | |
| "grad_norm": 0.5664693515511934, | |
| "learning_rate": 4.917030223798057e-06, | |
| "loss": 0.1339, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 4.9108280254777075, | |
| "grad_norm": 0.5684591845348423, | |
| "learning_rate": 4.834529604840686e-06, | |
| "loss": 0.1155, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 4.929936305732484, | |
| "grad_norm": 0.5981307471923145, | |
| "learning_rate": 4.7525057579209775e-06, | |
| "loss": 0.1278, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 4.949044585987261, | |
| "grad_norm": 0.6364067747914278, | |
| "learning_rate": 4.670966253826027e-06, | |
| "loss": 0.1066, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 4.968152866242038, | |
| "grad_norm": 0.5939598141476468, | |
| "learning_rate": 4.589918618638173e-06, | |
| "loss": 0.1439, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.987261146496815, | |
| "grad_norm": 0.6130361296769391, | |
| "learning_rate": 4.5093703330403385e-06, | |
| "loss": 0.1017, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 5.006369426751593, | |
| "grad_norm": 0.5008806887644277, | |
| "learning_rate": 4.429328831625565e-06, | |
| "loss": 0.091, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 5.025477707006369, | |
| "grad_norm": 0.4740521906355305, | |
| "learning_rate": 4.349801502210801e-06, | |
| "loss": 0.0822, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 5.044585987261146, | |
| "grad_norm": 0.44996258823609503, | |
| "learning_rate": 4.270795685155001e-06, | |
| "loss": 0.0657, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 5.063694267515924, | |
| "grad_norm": 0.4626707926362902, | |
| "learning_rate": 4.192318672681631e-06, | |
| "loss": 0.0712, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.082802547770701, | |
| "grad_norm": 0.4330055862509541, | |
| "learning_rate": 4.1143777082055715e-06, | |
| "loss": 0.0829, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 5.101910828025478, | |
| "grad_norm": 0.4598446407166053, | |
| "learning_rate": 4.036979985664566e-06, | |
| "loss": 0.1036, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 5.1210191082802545, | |
| "grad_norm": 0.49137420527897036, | |
| "learning_rate": 3.960132648855226e-06, | |
| "loss": 0.0819, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 5.140127388535032, | |
| "grad_norm": 0.4834112811966423, | |
| "learning_rate": 3.883842790773647e-06, | |
| "loss": 0.0862, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 5.159235668789809, | |
| "grad_norm": 0.555011867611463, | |
| "learning_rate": 3.8081174529607346e-06, | |
| "loss": 0.0826, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.178343949044586, | |
| "grad_norm": 0.45943211313271454, | |
| "learning_rate": 3.732963624852275e-06, | |
| "loss": 0.0694, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 5.197452229299363, | |
| "grad_norm": 0.4639094975421008, | |
| "learning_rate": 3.6583882431338047e-06, | |
| "loss": 0.0545, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 5.2165605095541405, | |
| "grad_norm": 0.46833755175037156, | |
| "learning_rate": 3.584398191100341e-06, | |
| "loss": 0.0836, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 5.235668789808917, | |
| "grad_norm": 0.5094975485065651, | |
| "learning_rate": 3.511000298021098e-06, | |
| "loss": 0.0844, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 5.254777070063694, | |
| "grad_norm": 0.4494664464442945, | |
| "learning_rate": 3.4382013385090985e-06, | |
| "loss": 0.1022, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.273885350318471, | |
| "grad_norm": 0.430587010997806, | |
| "learning_rate": 3.3660080318959043e-06, | |
| "loss": 0.0612, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 5.292993630573249, | |
| "grad_norm": 0.40568632150855466, | |
| "learning_rate": 3.2944270416114256e-06, | |
| "loss": 0.0637, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 5.312101910828026, | |
| "grad_norm": 0.4247687364924424, | |
| "learning_rate": 3.223464974568874e-06, | |
| "loss": 0.0951, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 5.3312101910828025, | |
| "grad_norm": 0.4203860396637708, | |
| "learning_rate": 3.153128380554941e-06, | |
| "loss": 0.0818, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 5.350318471337579, | |
| "grad_norm": 0.4183770518375645, | |
| "learning_rate": 3.0834237516252817e-06, | |
| "loss": 0.089, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.369426751592357, | |
| "grad_norm": 0.4058074699072587, | |
| "learning_rate": 3.0143575215052732e-06, | |
| "loss": 0.0836, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 5.388535031847134, | |
| "grad_norm": 0.3617226255237594, | |
| "learning_rate": 2.94593606499619e-06, | |
| "loss": 0.0485, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 5.407643312101911, | |
| "grad_norm": 0.4346354860683495, | |
| "learning_rate": 2.878165697386812e-06, | |
| "loss": 0.0899, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 5.426751592356688, | |
| "grad_norm": 0.4328661202373131, | |
| "learning_rate": 2.8110526738705345e-06, | |
| "loss": 0.0718, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 5.445859872611465, | |
| "grad_norm": 0.4463097565599843, | |
| "learning_rate": 2.7446031889679893e-06, | |
| "loss": 0.0826, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 5.464968152866242, | |
| "grad_norm": 0.4406773749162803, | |
| "learning_rate": 2.678823375955314e-06, | |
| "loss": 0.0685, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 5.484076433121019, | |
| "grad_norm": 0.4194123126293836, | |
| "learning_rate": 2.6137193062980506e-06, | |
| "loss": 0.0872, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 5.503184713375796, | |
| "grad_norm": 0.43930633536784436, | |
| "learning_rate": 2.5492969890907383e-06, | |
| "loss": 0.079, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 5.522292993630574, | |
| "grad_norm": 0.4429967810654036, | |
| "learning_rate": 2.485562370502279e-06, | |
| "loss": 0.0723, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 5.54140127388535, | |
| "grad_norm": 0.38961751873884726, | |
| "learning_rate": 2.4225213332271203e-06, | |
| "loss": 0.0828, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.560509554140127, | |
| "grad_norm": 0.4354990011032463, | |
| "learning_rate": 2.3601796959422585e-06, | |
| "loss": 0.0794, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 5.579617834394904, | |
| "grad_norm": 0.4399856065978314, | |
| "learning_rate": 2.2985432127701945e-06, | |
| "loss": 0.0947, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 5.598726114649682, | |
| "grad_norm": 0.39942606928684055, | |
| "learning_rate": 2.2376175727478346e-06, | |
| "loss": 0.0659, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 5.617834394904459, | |
| "grad_norm": 0.4226878330388677, | |
| "learning_rate": 2.1774083993013715e-06, | |
| "loss": 0.0973, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 5.6369426751592355, | |
| "grad_norm": 0.39260686788510807, | |
| "learning_rate": 2.1179212497272582e-06, | |
| "loss": 0.0782, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 5.656050955414012, | |
| "grad_norm": 0.36775177760897215, | |
| "learning_rate": 2.0591616146792705e-06, | |
| "loss": 0.0669, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 5.67515923566879, | |
| "grad_norm": 0.4699998862587698, | |
| "learning_rate": 2.0011349176617133e-06, | |
| "loss": 0.0727, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 5.694267515923567, | |
| "grad_norm": 0.38924128026976645, | |
| "learning_rate": 1.9438465145288377e-06, | |
| "loss": 0.0415, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 5.713375796178344, | |
| "grad_norm": 0.4047141514834865, | |
| "learning_rate": 1.8873016929904942e-06, | |
| "loss": 0.0731, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 5.732484076433121, | |
| "grad_norm": 0.42369205041349084, | |
| "learning_rate": 1.8315056721240831e-06, | |
| "loss": 0.0625, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.751592356687898, | |
| "grad_norm": 0.44037686412612786, | |
| "learning_rate": 1.7764636018928249e-06, | |
| "loss": 0.076, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 5.770700636942675, | |
| "grad_norm": 0.42701626585971403, | |
| "learning_rate": 1.722180562670428e-06, | |
| "loss": 0.1037, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 5.789808917197452, | |
| "grad_norm": 0.4321289371420163, | |
| "learning_rate": 1.6686615647721638e-06, | |
| "loss": 0.0589, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 5.80891719745223, | |
| "grad_norm": 0.4128682380834181, | |
| "learning_rate": 1.6159115479924259e-06, | |
| "loss": 0.0598, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 5.828025477707007, | |
| "grad_norm": 0.38040841468927883, | |
| "learning_rate": 1.5639353811487744e-06, | |
| "loss": 0.087, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 5.8471337579617835, | |
| "grad_norm": 0.386626285633781, | |
| "learning_rate": 1.5127378616325606e-06, | |
| "loss": 0.0626, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 5.86624203821656, | |
| "grad_norm": 0.4517222349712882, | |
| "learning_rate": 1.462323714966114e-06, | |
| "loss": 0.1057, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 5.885350318471337, | |
| "grad_norm": 0.3926830414947431, | |
| "learning_rate": 1.4126975943665844e-06, | |
| "loss": 0.0668, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 5.904458598726115, | |
| "grad_norm": 0.4323830177878288, | |
| "learning_rate": 1.3638640803164516e-06, | |
| "loss": 0.0801, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 5.923566878980892, | |
| "grad_norm": 0.4449523366792213, | |
| "learning_rate": 1.3158276801407432e-06, | |
| "loss": 0.0837, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.942675159235669, | |
| "grad_norm": 0.3777194865969392, | |
| "learning_rate": 1.2685928275910142e-06, | |
| "loss": 0.0867, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 5.961783439490446, | |
| "grad_norm": 0.40681186006426356, | |
| "learning_rate": 1.222163882436107e-06, | |
| "loss": 0.0769, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 5.980891719745223, | |
| "grad_norm": 0.4509886049646031, | |
| "learning_rate": 1.1765451300597574e-06, | |
| "loss": 0.072, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.3836657378760341, | |
| "learning_rate": 1.1317407810650372e-06, | |
| "loss": 0.0912, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 6.019108280254777, | |
| "grad_norm": 0.2897831911741826, | |
| "learning_rate": 1.0877549708857228e-06, | |
| "loss": 0.0582, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.038216560509555, | |
| "grad_norm": 0.36219579281581876, | |
| "learning_rate": 1.0445917594046073e-06, | |
| "loss": 0.095, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 6.057324840764331, | |
| "grad_norm": 0.3172624198549553, | |
| "learning_rate": 1.0022551305787564e-06, | |
| "loss": 0.0677, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 6.076433121019108, | |
| "grad_norm": 0.3328971252449723, | |
| "learning_rate": 9.607489920717983e-07, | |
| "loss": 0.0734, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 6.095541401273885, | |
| "grad_norm": 0.3462829070164449, | |
| "learning_rate": 9.200771748932513e-07, | |
| "loss": 0.0932, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 6.114649681528663, | |
| "grad_norm": 0.2979773377390498, | |
| "learning_rate": 8.802434330449128e-07, | |
| "loss": 0.0501, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.13375796178344, | |
| "grad_norm": 0.30877442111046816, | |
| "learning_rate": 8.412514431743657e-07, | |
| "loss": 0.0652, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 6.1528662420382165, | |
| "grad_norm": 0.2762953755167896, | |
| "learning_rate": 8.031048042356393e-07, | |
| "loss": 0.0611, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 6.171974522292993, | |
| "grad_norm": 0.31449200974883795, | |
| "learning_rate": 7.65807037157007e-07, | |
| "loss": 0.0672, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 6.191082802547771, | |
| "grad_norm": 0.3028008193328779, | |
| "learning_rate": 7.293615845160196e-07, | |
| "loss": 0.0709, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 6.210191082802548, | |
| "grad_norm": 0.29945853487475904, | |
| "learning_rate": 6.937718102217461e-07, | |
| "loss": 0.0539, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.229299363057325, | |
| "grad_norm": 0.2995993442678918, | |
| "learning_rate": 6.590409992042957e-07, | |
| "loss": 0.0537, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 6.248407643312102, | |
| "grad_norm": 0.2976107306131544, | |
| "learning_rate": 6.251723571116031e-07, | |
| "loss": 0.0602, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 6.267515923566879, | |
| "grad_norm": 0.34723393391763074, | |
| "learning_rate": 5.921690100135713e-07, | |
| "loss": 0.0566, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 6.286624203821656, | |
| "grad_norm": 0.29267196212983465, | |
| "learning_rate": 5.600340041135133e-07, | |
| "loss": 0.0561, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 6.305732484076433, | |
| "grad_norm": 0.30989443431520364, | |
| "learning_rate": 5.287703054670012e-07, | |
| "loss": 0.0868, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 6.32484076433121, | |
| "grad_norm": 0.3452062246224452, | |
| "learning_rate": 4.983807997080925e-07, | |
| "loss": 0.0774, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 6.343949044585988, | |
| "grad_norm": 0.2991123427207596, | |
| "learning_rate": 4.6886829178299676e-07, | |
| "loss": 0.0638, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 6.3630573248407645, | |
| "grad_norm": 0.251877453202914, | |
| "learning_rate": 4.402355056911656e-07, | |
| "loss": 0.0416, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 6.382165605095541, | |
| "grad_norm": 0.26908451663125665, | |
| "learning_rate": 4.124850842338779e-07, | |
| "loss": 0.0484, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 6.401273885350318, | |
| "grad_norm": 0.34363749925099824, | |
| "learning_rate": 3.8561958877030957e-07, | |
| "loss": 0.071, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 6.420382165605096, | |
| "grad_norm": 0.29923360300918206, | |
| "learning_rate": 3.5964149898111587e-07, | |
| "loss": 0.0415, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 6.439490445859873, | |
| "grad_norm": 0.30645214293590295, | |
| "learning_rate": 3.345532126395579e-07, | |
| "loss": 0.0728, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 6.45859872611465, | |
| "grad_norm": 0.37628383262174114, | |
| "learning_rate": 3.1035704539019384e-07, | |
| "loss": 0.0729, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 6.477707006369426, | |
| "grad_norm": 0.3034257258064961, | |
| "learning_rate": 2.870552305351382e-07, | |
| "loss": 0.0606, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 6.496815286624204, | |
| "grad_norm": 0.31100043418683965, | |
| "learning_rate": 2.646499188279328e-07, | |
| "loss": 0.0775, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 6.515923566878981, | |
| "grad_norm": 0.2898244735717334, | |
| "learning_rate": 2.4314317827503375e-07, | |
| "loss": 0.0542, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 6.535031847133758, | |
| "grad_norm": 0.2744711975330943, | |
| "learning_rate": 2.2253699394493066e-07, | |
| "loss": 0.0463, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 6.554140127388535, | |
| "grad_norm": 0.3295027537161378, | |
| "learning_rate": 2.028332677849254e-07, | |
| "loss": 0.0587, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 6.573248407643312, | |
| "grad_norm": 0.2739865340285092, | |
| "learning_rate": 1.840338184455881e-07, | |
| "loss": 0.0621, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 6.592356687898089, | |
| "grad_norm": 0.3049696335907913, | |
| "learning_rate": 1.6614038111289034e-07, | |
| "loss": 0.0454, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 6.611464968152866, | |
| "grad_norm": 0.30860070843186094, | |
| "learning_rate": 1.49154607348051e-07, | |
| "loss": 0.065, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 6.630573248407643, | |
| "grad_norm": 0.3280703997948172, | |
| "learning_rate": 1.330780649350938e-07, | |
| "loss": 0.0564, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 6.649681528662421, | |
| "grad_norm": 0.34587757692766835, | |
| "learning_rate": 1.1791223773614635e-07, | |
| "loss": 0.0718, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 6.6687898089171975, | |
| "grad_norm": 0.3092259328139762, | |
| "learning_rate": 1.0365852555447642e-07, | |
| "loss": 0.057, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 6.687898089171974, | |
| "grad_norm": 0.31877342074996656, | |
| "learning_rate": 9.031824400528854e-08, | |
| "loss": 0.0652, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.707006369426751, | |
| "grad_norm": 0.2960124524937976, | |
| "learning_rate": 7.789262439430012e-08, | |
| "loss": 0.0651, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 6.726114649681529, | |
| "grad_norm": 0.3243223624677724, | |
| "learning_rate": 6.638281360408339e-08, | |
| "loss": 0.0831, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 6.745222929936306, | |
| "grad_norm": 0.3050317883424216, | |
| "learning_rate": 5.578987398821345e-08, | |
| "loss": 0.0525, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 6.764331210191083, | |
| "grad_norm": 0.2926711184299434, | |
| "learning_rate": 4.6114783273213395e-08, | |
| "loss": 0.0722, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 6.7834394904458595, | |
| "grad_norm": 0.3389636588784217, | |
| "learning_rate": 3.735843446830867e-08, | |
| "loss": 0.076, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 6.802547770700637, | |
| "grad_norm": 0.3193710836037101, | |
| "learning_rate": 2.9521635783001932e-08, | |
| "loss": 0.0478, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 6.821656050955414, | |
| "grad_norm": 0.28200675616323173, | |
| "learning_rate": 2.2605110552477162e-08, | |
| "loss": 0.0553, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 6.840764331210191, | |
| "grad_norm": 0.27640347185235314, | |
| "learning_rate": 1.6609497170834154e-08, | |
| "loss": 0.0503, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 6.859872611464969, | |
| "grad_norm": 0.3047895437024314, | |
| "learning_rate": 1.1535349032167908e-08, | |
| "loss": 0.0698, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 6.8789808917197455, | |
| "grad_norm": 0.2673494245873522, | |
| "learning_rate": 7.3831344794872415e-09, | |
| "loss": 0.0591, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 6.898089171974522, | |
| "grad_norm": 0.2898489566203137, | |
| "learning_rate": 4.153236761488266e-09, | |
| "loss": 0.0523, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 6.917197452229299, | |
| "grad_norm": 0.3143330666373118, | |
| "learning_rate": 1.8459539971804608e-09, | |
| "loss": 0.0608, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 6.936305732484076, | |
| "grad_norm": 0.30228762531335357, | |
| "learning_rate": 4.614991483686826e-10, | |
| "loss": 0.0692, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 6.955414012738854, | |
| "grad_norm": 0.3087509415829173, | |
| "learning_rate": 0.0, | |
| "loss": 0.0518, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 6.955414012738854, | |
| "step": 364, | |
| "total_flos": 8.151821126736282e+16, | |
| "train_loss": 0.35076490221647444, | |
| "train_runtime": 2395.9288, | |
| "train_samples_per_second": 14.605, | |
| "train_steps_per_second": 0.152 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 364, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.151821126736282e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |