| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.975629569455727, | |
| "eval_steps": 500, | |
| "global_step": 380, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012997562956945572, | |
| "grad_norm": 6.099751379930322, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 1.0219, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.025995125913891144, | |
| "grad_norm": 6.068929688233013, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 1.0148, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03899268887083672, | |
| "grad_norm": 5.607482856386817, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 0.9978, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05199025182778229, | |
| "grad_norm": 4.0344324115028956, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 0.9695, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06498781478472786, | |
| "grad_norm": 2.3652828088294218, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 0.9321, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07798537774167344, | |
| "grad_norm": 4.801199261631304, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 0.9538, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09098294069861901, | |
| "grad_norm": 4.831210221526376, | |
| "learning_rate": 1.4736842105263159e-05, | |
| "loss": 0.9455, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.10398050365556458, | |
| "grad_norm": 5.705571120150557, | |
| "learning_rate": 1.6842105263157896e-05, | |
| "loss": 0.9346, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11697806661251016, | |
| "grad_norm": 4.118034247132142, | |
| "learning_rate": 1.894736842105263e-05, | |
| "loss": 0.9103, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.12997562956945571, | |
| "grad_norm": 2.7464416924594333, | |
| "learning_rate": 2.105263157894737e-05, | |
| "loss": 0.8666, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1429731925264013, | |
| "grad_norm": 1.8112884281407549, | |
| "learning_rate": 2.3157894736842107e-05, | |
| "loss": 0.8288, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15597075548334688, | |
| "grad_norm": 1.6168217780438021, | |
| "learning_rate": 2.526315789473684e-05, | |
| "loss": 0.8095, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.16896831844029245, | |
| "grad_norm": 1.0800910213719421, | |
| "learning_rate": 2.7368421052631583e-05, | |
| "loss": 0.7876, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.18196588139723802, | |
| "grad_norm": 1.273865117132841, | |
| "learning_rate": 2.9473684210526317e-05, | |
| "loss": 0.7801, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19496344435418358, | |
| "grad_norm": 1.4101116134429739, | |
| "learning_rate": 3.157894736842106e-05, | |
| "loss": 0.7769, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.20796100731112915, | |
| "grad_norm": 1.3300011416464073, | |
| "learning_rate": 3.368421052631579e-05, | |
| "loss": 0.7705, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.22095857026807472, | |
| "grad_norm": 1.2303404991910958, | |
| "learning_rate": 3.578947368421053e-05, | |
| "loss": 0.7457, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23395613322502032, | |
| "grad_norm": 1.2600947996998637, | |
| "learning_rate": 3.789473684210526e-05, | |
| "loss": 0.7414, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2469536961819659, | |
| "grad_norm": 0.9415351295551974, | |
| "learning_rate": 4e-05, | |
| "loss": 0.731, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.25995125913891143, | |
| "grad_norm": 1.6378942120476991, | |
| "learning_rate": 4.210526315789474e-05, | |
| "loss": 0.7402, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.272948822095857, | |
| "grad_norm": 1.452156312513347, | |
| "learning_rate": 4.421052631578948e-05, | |
| "loss": 0.7293, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2859463850528026, | |
| "grad_norm": 1.1024110144705295, | |
| "learning_rate": 4.6315789473684214e-05, | |
| "loss": 0.7143, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.29894394800974816, | |
| "grad_norm": 1.9887153689227788, | |
| "learning_rate": 4.842105263157895e-05, | |
| "loss": 0.7255, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.31194151096669376, | |
| "grad_norm": 1.1903531007119599, | |
| "learning_rate": 5.052631578947368e-05, | |
| "loss": 0.7177, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.3249390739236393, | |
| "grad_norm": 1.9904927095675595, | |
| "learning_rate": 5.263157894736843e-05, | |
| "loss": 0.7215, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3379366368805849, | |
| "grad_norm": 1.6681859777545944, | |
| "learning_rate": 5.4736842105263165e-05, | |
| "loss": 0.7009, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.35093419983753044, | |
| "grad_norm": 1.6209941695574477, | |
| "learning_rate": 5.68421052631579e-05, | |
| "loss": 0.7052, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.36393176279447603, | |
| "grad_norm": 1.98849797370751, | |
| "learning_rate": 5.8947368421052634e-05, | |
| "loss": 0.7004, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.37692932575142163, | |
| "grad_norm": 1.12064775343047, | |
| "learning_rate": 6.105263157894738e-05, | |
| "loss": 0.6898, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.38992688870836717, | |
| "grad_norm": 2.2171415848887475, | |
| "learning_rate": 6.315789473684212e-05, | |
| "loss": 0.6866, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.40292445166531277, | |
| "grad_norm": 1.5015021280904342, | |
| "learning_rate": 6.526315789473685e-05, | |
| "loss": 0.6899, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4159220146222583, | |
| "grad_norm": 2.1943095640217583, | |
| "learning_rate": 6.736842105263159e-05, | |
| "loss": 0.6852, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4289195775792039, | |
| "grad_norm": 1.7052014979491807, | |
| "learning_rate": 6.947368421052632e-05, | |
| "loss": 0.6898, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.44191714053614944, | |
| "grad_norm": 1.4787351390255603, | |
| "learning_rate": 7.157894736842105e-05, | |
| "loss": 0.6812, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.45491470349309504, | |
| "grad_norm": 1.1982328674577414, | |
| "learning_rate": 7.368421052631579e-05, | |
| "loss": 0.6827, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.46791226645004064, | |
| "grad_norm": 2.055421861681658, | |
| "learning_rate": 7.578947368421052e-05, | |
| "loss": 0.6747, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4809098294069862, | |
| "grad_norm": 1.3504861459840583, | |
| "learning_rate": 7.789473684210527e-05, | |
| "loss": 0.6697, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4939073923639318, | |
| "grad_norm": 1.9872016361599776, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6748, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5069049553208773, | |
| "grad_norm": 1.9190088069026263, | |
| "learning_rate": 7.99983123807325e-05, | |
| "loss": 0.6721, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5199025182778229, | |
| "grad_norm": 1.265154939225945, | |
| "learning_rate": 7.999324966533291e-05, | |
| "loss": 0.6603, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5329000812347685, | |
| "grad_norm": 3.2164002414551978, | |
| "learning_rate": 7.998481228099806e-05, | |
| "loss": 0.6723, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.545897644191714, | |
| "grad_norm": 2.571438671486186, | |
| "learning_rate": 7.997300093968255e-05, | |
| "loss": 0.6806, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5588952071486596, | |
| "grad_norm": 1.8156950022852132, | |
| "learning_rate": 7.995781663803876e-05, | |
| "loss": 0.6635, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5718927701056052, | |
| "grad_norm": 1.3254069741419403, | |
| "learning_rate": 7.993926065733265e-05, | |
| "loss": 0.6637, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5848903330625508, | |
| "grad_norm": 0.9530737655919242, | |
| "learning_rate": 7.991733456333579e-05, | |
| "loss": 0.6636, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5978878960194963, | |
| "grad_norm": 1.3664556141065045, | |
| "learning_rate": 7.98920402061931e-05, | |
| "loss": 0.6616, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6108854589764419, | |
| "grad_norm": 1.2839159067727302, | |
| "learning_rate": 7.98633797202668e-05, | |
| "loss": 0.6656, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6238830219333875, | |
| "grad_norm": 1.1439614314511357, | |
| "learning_rate": 7.98313555239563e-05, | |
| "loss": 0.6639, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6368805848903331, | |
| "grad_norm": 1.8097099672517978, | |
| "learning_rate": 7.979597031949415e-05, | |
| "loss": 0.6788, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6498781478472786, | |
| "grad_norm": 1.0800803816613773, | |
| "learning_rate": 7.975722709271799e-05, | |
| "loss": 0.6578, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6628757108042242, | |
| "grad_norm": 2.014427030879868, | |
| "learning_rate": 7.97151291128186e-05, | |
| "loss": 0.664, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6758732737611698, | |
| "grad_norm": 1.5251902643366386, | |
| "learning_rate": 7.96696799320641e-05, | |
| "loss": 0.6617, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6888708367181153, | |
| "grad_norm": 1.4609799533681953, | |
| "learning_rate": 7.962088338550013e-05, | |
| "loss": 0.6544, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7018683996750609, | |
| "grad_norm": 1.2857979651137834, | |
| "learning_rate": 7.956874359062632e-05, | |
| "loss": 0.6563, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.7148659626320065, | |
| "grad_norm": 1.167851324111819, | |
| "learning_rate": 7.951326494704878e-05, | |
| "loss": 0.6432, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7278635255889521, | |
| "grad_norm": 0.923367616793914, | |
| "learning_rate": 7.94544521361089e-05, | |
| "loss": 0.6458, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7408610885458976, | |
| "grad_norm": 1.2536213795566458, | |
| "learning_rate": 7.939231012048833e-05, | |
| "loss": 0.6483, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7538586515028433, | |
| "grad_norm": 1.2119787729723712, | |
| "learning_rate": 7.932684414379021e-05, | |
| "loss": 0.6461, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7668562144597888, | |
| "grad_norm": 0.7624960645952309, | |
| "learning_rate": 7.925805973009672e-05, | |
| "loss": 0.6347, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7798537774167343, | |
| "grad_norm": 0.9917480604742027, | |
| "learning_rate": 7.918596268350296e-05, | |
| "loss": 0.639, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7928513403736799, | |
| "grad_norm": 1.3671397849973843, | |
| "learning_rate": 7.911055908762718e-05, | |
| "loss": 0.6416, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.8058489033306255, | |
| "grad_norm": 0.7823232773420852, | |
| "learning_rate": 7.903185530509743e-05, | |
| "loss": 0.633, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8188464662875711, | |
| "grad_norm": 0.8711197066128451, | |
| "learning_rate": 7.894985797701472e-05, | |
| "loss": 0.6348, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8318440292445166, | |
| "grad_norm": 1.0138573543907548, | |
| "learning_rate": 7.886457402239256e-05, | |
| "loss": 0.6317, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8448415922014623, | |
| "grad_norm": 1.328902077693916, | |
| "learning_rate": 7.877601063757323e-05, | |
| "loss": 0.641, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8578391551584078, | |
| "grad_norm": 0.8975115888522265, | |
| "learning_rate": 7.868417529562043e-05, | |
| "loss": 0.6225, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8708367181153533, | |
| "grad_norm": 0.9433279422872343, | |
| "learning_rate": 7.858907574568882e-05, | |
| "loss": 0.6318, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8838342810722989, | |
| "grad_norm": 0.9051780877745009, | |
| "learning_rate": 7.849072001237001e-05, | |
| "loss": 0.623, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8968318440292445, | |
| "grad_norm": 1.1238293364288126, | |
| "learning_rate": 7.838911639501557e-05, | |
| "loss": 0.6197, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.9098294069861901, | |
| "grad_norm": 0.9616875737950008, | |
| "learning_rate": 7.828427346703657e-05, | |
| "loss": 0.6151, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9228269699431356, | |
| "grad_norm": 0.7925602555435675, | |
| "learning_rate": 7.81762000751803e-05, | |
| "loss": 0.6204, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9358245329000813, | |
| "grad_norm": 0.8202368746016689, | |
| "learning_rate": 7.806490533878368e-05, | |
| "loss": 0.6206, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9488220958570268, | |
| "grad_norm": 0.7281381020297469, | |
| "learning_rate": 7.795039864900378e-05, | |
| "loss": 0.6145, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9618196588139724, | |
| "grad_norm": 0.951203457585769, | |
| "learning_rate": 7.783268966802539e-05, | |
| "loss": 0.6271, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.974817221770918, | |
| "grad_norm": 1.079648600019301, | |
| "learning_rate": 7.771178832824573e-05, | |
| "loss": 0.6142, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9878147847278635, | |
| "grad_norm": 0.7728669124672748, | |
| "learning_rate": 7.758770483143634e-05, | |
| "loss": 0.6123, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.0024370430544274, | |
| "grad_norm": 0.9574347690194315, | |
| "learning_rate": 7.74604496478822e-05, | |
| "loss": 0.615, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.0154346060113728, | |
| "grad_norm": 1.0135972450552604, | |
| "learning_rate": 7.733003351549829e-05, | |
| "loss": 0.5888, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0284321689683185, | |
| "grad_norm": 1.1758211244717183, | |
| "learning_rate": 7.719646743892352e-05, | |
| "loss": 0.5932, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.0414297319252641, | |
| "grad_norm": 1.0448938050043781, | |
| "learning_rate": 7.705976268859207e-05, | |
| "loss": 0.5934, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0544272948822095, | |
| "grad_norm": 0.9104488344215643, | |
| "learning_rate": 7.691993079978252e-05, | |
| "loss": 0.5815, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0674248578391552, | |
| "grad_norm": 0.7412780882809253, | |
| "learning_rate": 7.677698357164431e-05, | |
| "loss": 0.5752, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0804224207961006, | |
| "grad_norm": 0.7417024796170562, | |
| "learning_rate": 7.663093306620231e-05, | |
| "loss": 0.5792, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0934199837530463, | |
| "grad_norm": 0.7190010808453768, | |
| "learning_rate": 7.648179160733883e-05, | |
| "loss": 0.5753, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.106417546709992, | |
| "grad_norm": 0.7427027225824168, | |
| "learning_rate": 7.632957177975387e-05, | |
| "loss": 0.5777, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1194151096669374, | |
| "grad_norm": 0.9295269491982405, | |
| "learning_rate": 7.61742864279031e-05, | |
| "loss": 0.5834, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.132412672623883, | |
| "grad_norm": 0.9634417055121992, | |
| "learning_rate": 7.601594865491414e-05, | |
| "loss": 0.5742, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1454102355808287, | |
| "grad_norm": 0.9476309894224839, | |
| "learning_rate": 7.585457182148081e-05, | |
| "loss": 0.5784, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.158407798537774, | |
| "grad_norm": 0.9101385250862156, | |
| "learning_rate": 7.569016954473577e-05, | |
| "loss": 0.5826, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.1714053614947197, | |
| "grad_norm": 0.9143144480170803, | |
| "learning_rate": 7.552275569710152e-05, | |
| "loss": 0.5852, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1844029244516654, | |
| "grad_norm": 0.6885600525748597, | |
| "learning_rate": 7.535234440511979e-05, | |
| "loss": 0.5763, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1974004874086108, | |
| "grad_norm": 0.5214486371121746, | |
| "learning_rate": 7.517895004825956e-05, | |
| "loss": 0.5787, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.2103980503655565, | |
| "grad_norm": 0.758408482295764, | |
| "learning_rate": 7.500258725770375e-05, | |
| "loss": 0.5746, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2233956133225021, | |
| "grad_norm": 0.7125338265833929, | |
| "learning_rate": 7.48232709151145e-05, | |
| "loss": 0.5703, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.2363931762794476, | |
| "grad_norm": 0.6377296028835163, | |
| "learning_rate": 7.464101615137756e-05, | |
| "loss": 0.579, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2493907392363932, | |
| "grad_norm": 0.5870416970514215, | |
| "learning_rate": 7.445583834532546e-05, | |
| "loss": 0.5637, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2623883021933389, | |
| "grad_norm": 0.4860633167448116, | |
| "learning_rate": 7.426775312243986e-05, | |
| "loss": 0.5756, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2753858651502843, | |
| "grad_norm": 0.5083572483496955, | |
| "learning_rate": 7.407677635353308e-05, | |
| "loss": 0.5734, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.28838342810723, | |
| "grad_norm": 0.5305435155519614, | |
| "learning_rate": 7.388292415340888e-05, | |
| "loss": 0.5682, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.3013809910641756, | |
| "grad_norm": 0.4176759940201937, | |
| "learning_rate": 7.368621287950264e-05, | |
| "loss": 0.5643, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.314378554021121, | |
| "grad_norm": 0.4583116745957642, | |
| "learning_rate": 7.348665913050115e-05, | |
| "loss": 0.5602, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3273761169780667, | |
| "grad_norm": 0.5426624427041704, | |
| "learning_rate": 7.328427974494201e-05, | |
| "loss": 0.5672, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.340373679935012, | |
| "grad_norm": 0.4189086182858418, | |
| "learning_rate": 7.307909179979274e-05, | |
| "loss": 0.5648, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3533712428919578, | |
| "grad_norm": 0.39949243757403974, | |
| "learning_rate": 7.28711126090098e-05, | |
| "loss": 0.566, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3663688058489034, | |
| "grad_norm": 0.43131027290320334, | |
| "learning_rate": 7.266035972207773e-05, | |
| "loss": 0.5571, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3793663688058488, | |
| "grad_norm": 0.40582965588538134, | |
| "learning_rate": 7.24468509225281e-05, | |
| "loss": 0.5693, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3923639317627945, | |
| "grad_norm": 0.4761372341291519, | |
| "learning_rate": 7.223060422643914e-05, | |
| "loss": 0.5686, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.40536149471974, | |
| "grad_norm": 0.5727339140771481, | |
| "learning_rate": 7.201163788091536e-05, | |
| "loss": 0.5678, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.4183590576766856, | |
| "grad_norm": 0.7085163969661245, | |
| "learning_rate": 7.178997036254799e-05, | |
| "loss": 0.5683, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.4313566206336312, | |
| "grad_norm": 0.80819559722067, | |
| "learning_rate": 7.156562037585576e-05, | |
| "loss": 0.5665, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4443541835905767, | |
| "grad_norm": 0.793253904366816, | |
| "learning_rate": 7.133860685170665e-05, | |
| "loss": 0.5745, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4573517465475223, | |
| "grad_norm": 0.7256059178824779, | |
| "learning_rate": 7.110894894572056e-05, | |
| "loss": 0.5645, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.470349309504468, | |
| "grad_norm": 0.5014971965016265, | |
| "learning_rate": 7.087666603665284e-05, | |
| "loss": 0.5606, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4833468724614134, | |
| "grad_norm": 0.43914912895871583, | |
| "learning_rate": 7.064177772475912e-05, | |
| "loss": 0.5674, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.496344435418359, | |
| "grad_norm": 0.4961335507396086, | |
| "learning_rate": 7.040430383014146e-05, | |
| "loss": 0.5655, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.5093419983753047, | |
| "grad_norm": 0.5185842695541142, | |
| "learning_rate": 7.016426439107586e-05, | |
| "loss": 0.5653, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.5223395613322501, | |
| "grad_norm": 0.37245930997617116, | |
| "learning_rate": 6.992167966232143e-05, | |
| "loss": 0.5567, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5353371242891958, | |
| "grad_norm": 0.5310947508784376, | |
| "learning_rate": 6.967657011341126e-05, | |
| "loss": 0.575, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5483346872461414, | |
| "grad_norm": 0.7628101414612303, | |
| "learning_rate": 6.942895642692527e-05, | |
| "loss": 0.5624, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.5613322502030869, | |
| "grad_norm": 0.7910046561291942, | |
| "learning_rate": 6.917885949674483e-05, | |
| "loss": 0.5667, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5743298131600325, | |
| "grad_norm": 0.7246609366699134, | |
| "learning_rate": 6.892630042628988e-05, | |
| "loss": 0.557, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5873273761169782, | |
| "grad_norm": 0.6852631021753094, | |
| "learning_rate": 6.867130052673806e-05, | |
| "loss": 0.5638, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.6003249390739236, | |
| "grad_norm": 0.6947100326982237, | |
| "learning_rate": 6.841388131522656e-05, | |
| "loss": 0.5626, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.6133225020308692, | |
| "grad_norm": 0.5891017692387462, | |
| "learning_rate": 6.815406451303647e-05, | |
| "loss": 0.5632, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.626320064987815, | |
| "grad_norm": 0.44282420651875987, | |
| "learning_rate": 6.789187204375981e-05, | |
| "loss": 0.5643, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6393176279447603, | |
| "grad_norm": 0.3540188196234301, | |
| "learning_rate": 6.762732603144978e-05, | |
| "loss": 0.5611, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.652315190901706, | |
| "grad_norm": 0.38606767224591043, | |
| "learning_rate": 6.736044879875373e-05, | |
| "loss": 0.5548, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6653127538586516, | |
| "grad_norm": 0.4847276958422751, | |
| "learning_rate": 6.709126286502965e-05, | |
| "loss": 0.5634, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.678310316815597, | |
| "grad_norm": 0.48098159010027874, | |
| "learning_rate": 6.681979094444596e-05, | |
| "loss": 0.558, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6913078797725425, | |
| "grad_norm": 0.35223251057931265, | |
| "learning_rate": 6.654605594406486e-05, | |
| "loss": 0.5584, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.7043054427294884, | |
| "grad_norm": 0.3582278064607583, | |
| "learning_rate": 6.627008096190938e-05, | |
| "loss": 0.5581, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.7173030056864338, | |
| "grad_norm": 0.4148482135356544, | |
| "learning_rate": 6.59918892850144e-05, | |
| "loss": 0.5604, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7303005686433792, | |
| "grad_norm": 0.5160390925469622, | |
| "learning_rate": 6.571150438746157e-05, | |
| "loss": 0.5578, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.743298131600325, | |
| "grad_norm": 0.6691571622471794, | |
| "learning_rate": 6.542894992839873e-05, | |
| "loss": 0.5521, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.7562956945572705, | |
| "grad_norm": 0.7398206439178091, | |
| "learning_rate": 6.514424975004329e-05, | |
| "loss": 0.5559, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.769293257514216, | |
| "grad_norm": 0.6904604759405913, | |
| "learning_rate": 6.48574278756706e-05, | |
| "loss": 0.556, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7822908204711616, | |
| "grad_norm": 0.5927997707542059, | |
| "learning_rate": 6.456850850758673e-05, | |
| "loss": 0.5511, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7952883834281073, | |
| "grad_norm": 0.598576202040419, | |
| "learning_rate": 6.427751602508628e-05, | |
| "loss": 0.5517, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.8082859463850527, | |
| "grad_norm": 0.5752299054086354, | |
| "learning_rate": 6.398447498239527e-05, | |
| "loss": 0.5603, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.8212835093419983, | |
| "grad_norm": 0.47947265005962353, | |
| "learning_rate": 6.368941010659921e-05, | |
| "loss": 0.5583, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.834281072298944, | |
| "grad_norm": 0.3589925554258543, | |
| "learning_rate": 6.339234629555655e-05, | |
| "loss": 0.5566, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8472786352558894, | |
| "grad_norm": 0.3418394937521007, | |
| "learning_rate": 6.309330861579786e-05, | |
| "loss": 0.5592, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.860276198212835, | |
| "grad_norm": 0.4539321991211633, | |
| "learning_rate": 6.279232230041065e-05, | |
| "loss": 0.5544, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8732737611697807, | |
| "grad_norm": 0.44216438200356556, | |
| "learning_rate": 6.248941274691017e-05, | |
| "loss": 0.5519, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8862713241267262, | |
| "grad_norm": 0.3237161862715092, | |
| "learning_rate": 6.218460551509636e-05, | |
| "loss": 0.5471, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8992688870836718, | |
| "grad_norm": 0.3051484675988478, | |
| "learning_rate": 6.18779263248971e-05, | |
| "loss": 0.5551, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.9122664500406175, | |
| "grad_norm": 0.459140656732003, | |
| "learning_rate": 6.156940105419785e-05, | |
| "loss": 0.5491, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.925264012997563, | |
| "grad_norm": 0.5335980534072164, | |
| "learning_rate": 6.125905573665824e-05, | |
| "loss": 0.5515, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9382615759545085, | |
| "grad_norm": 0.4199748973939331, | |
| "learning_rate": 6.094691655951512e-05, | |
| "loss": 0.554, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.9512591389114542, | |
| "grad_norm": 0.37821939152611067, | |
| "learning_rate": 6.063300986137297e-05, | |
| "loss": 0.5515, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9642567018683996, | |
| "grad_norm": 0.4525799105101174, | |
| "learning_rate": 6.0317362129981375e-05, | |
| "loss": 0.5514, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9772542648253453, | |
| "grad_norm": 0.4866209265488796, | |
| "learning_rate": 6.000000000000001e-05, | |
| "loss": 0.5557, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.990251827782291, | |
| "grad_norm": 0.4881518290915675, | |
| "learning_rate": 5.968095025075114e-05, | |
| "loss": 0.5581, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.0048740861088548, | |
| "grad_norm": 0.3855836959947548, | |
| "learning_rate": 5.936023980395997e-05, | |
| "loss": 0.5411, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.0178716490658, | |
| "grad_norm": 0.3731193378171015, | |
| "learning_rate": 5.903789572148295e-05, | |
| "loss": 0.5115, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.0308692120227456, | |
| "grad_norm": 0.4760587483033507, | |
| "learning_rate": 5.871394520302432e-05, | |
| "loss": 0.5036, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0438667749796915, | |
| "grad_norm": 0.5788173180313999, | |
| "learning_rate": 5.838841558384091e-05, | |
| "loss": 0.5023, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.056864337936637, | |
| "grad_norm": 0.610294677890403, | |
| "learning_rate": 5.806133433243558e-05, | |
| "loss": 0.5115, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0698619008935824, | |
| "grad_norm": 0.5785376971559186, | |
| "learning_rate": 5.7732729048239444e-05, | |
| "loss": 0.5059, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.0828594638505282, | |
| "grad_norm": 0.6361199902732823, | |
| "learning_rate": 5.740262745928293e-05, | |
| "loss": 0.5094, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0958570268074737, | |
| "grad_norm": 0.7055783417389399, | |
| "learning_rate": 5.707105741985615e-05, | |
| "loss": 0.5099, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.108854589764419, | |
| "grad_norm": 0.7551674489031295, | |
| "learning_rate": 5.673804690815845e-05, | |
| "loss": 0.5115, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.121852152721365, | |
| "grad_norm": 0.7239995903013289, | |
| "learning_rate": 5.6403624023937614e-05, | |
| "loss": 0.5031, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.1348497156783104, | |
| "grad_norm": 0.509463011523919, | |
| "learning_rate": 5.606781698611879e-05, | |
| "loss": 0.5021, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.147847278635256, | |
| "grad_norm": 0.4210575883324648, | |
| "learning_rate": 5.573065413042333e-05, | |
| "loss": 0.5053, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1608448415922012, | |
| "grad_norm": 0.5009273026595834, | |
| "learning_rate": 5.5392163906977835e-05, | |
| "loss": 0.507, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.173842404549147, | |
| "grad_norm": 0.47375681034352735, | |
| "learning_rate": 5.505237487791343e-05, | |
| "loss": 0.5042, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1868399675060926, | |
| "grad_norm": 0.37667913409010934, | |
| "learning_rate": 5.471131571495574e-05, | |
| "loss": 0.5006, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.199837530463038, | |
| "grad_norm": 0.292819297582803, | |
| "learning_rate": 5.4369015197005506e-05, | |
| "loss": 0.5102, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.212835093419984, | |
| "grad_norm": 0.4148005226049645, | |
| "learning_rate": 5.4025502207710184e-05, | |
| "loss": 0.5061, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.2258326563769293, | |
| "grad_norm": 0.34689504337022264, | |
| "learning_rate": 5.368080573302676e-05, | |
| "loss": 0.5016, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2388302193338747, | |
| "grad_norm": 0.3094307656424445, | |
| "learning_rate": 5.333495485877583e-05, | |
| "loss": 0.5014, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2518277822908206, | |
| "grad_norm": 0.3739855090857637, | |
| "learning_rate": 5.298797876818735e-05, | |
| "loss": 0.501, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.264825345247766, | |
| "grad_norm": 0.3427237300205951, | |
| "learning_rate": 5.263990673943811e-05, | |
| "loss": 0.5022, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.2778229082047114, | |
| "grad_norm": 0.2853871717746141, | |
| "learning_rate": 5.229076814318122e-05, | |
| "loss": 0.5024, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2908204711616573, | |
| "grad_norm": 0.30137576187728127, | |
| "learning_rate": 5.194059244006779e-05, | |
| "loss": 0.4956, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.3038180341186028, | |
| "grad_norm": 0.34898425966594765, | |
| "learning_rate": 5.158940917826099e-05, | |
| "loss": 0.5048, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.316815597075548, | |
| "grad_norm": 0.3375764565690276, | |
| "learning_rate": 5.123724799094279e-05, | |
| "loss": 0.4971, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.329813160032494, | |
| "grad_norm": 0.3526352081558193, | |
| "learning_rate": 5.088413859381341e-05, | |
| "loss": 0.4953, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.3428107229894395, | |
| "grad_norm": 0.2669067122173656, | |
| "learning_rate": 5.053011078258397e-05, | |
| "loss": 0.4988, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.355808285946385, | |
| "grad_norm": 0.2634804890652626, | |
| "learning_rate": 5.017519443046226e-05, | |
| "loss": 0.5062, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.368805848903331, | |
| "grad_norm": 0.3884943325282365, | |
| "learning_rate": 4.981941948563197e-05, | |
| "loss": 0.4947, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.381803411860276, | |
| "grad_norm": 0.30824639305661866, | |
| "learning_rate": 4.94628159687257e-05, | |
| "loss": 0.4959, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3948009748172217, | |
| "grad_norm": 0.2779177232068281, | |
| "learning_rate": 4.9105413970291747e-05, | |
| "loss": 0.5004, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.4077985377741675, | |
| "grad_norm": 0.2542309798243027, | |
| "learning_rate": 4.874724364825504e-05, | |
| "loss": 0.5059, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.420796100731113, | |
| "grad_norm": 0.2377609454443894, | |
| "learning_rate": 4.8388335225372416e-05, | |
| "loss": 0.5047, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.4337936636880584, | |
| "grad_norm": 0.2783658709624058, | |
| "learning_rate": 4.802871898668237e-05, | |
| "loss": 0.5019, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4467912266450043, | |
| "grad_norm": 0.212526701822668, | |
| "learning_rate": 4.7668425276949546e-05, | |
| "loss": 0.5055, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4597887896019497, | |
| "grad_norm": 0.23513126552745808, | |
| "learning_rate": 4.730748449810429e-05, | |
| "loss": 0.4955, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.472786352558895, | |
| "grad_norm": 0.24679411263913667, | |
| "learning_rate": 4.694592710667723e-05, | |
| "loss": 0.4954, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4857839155158405, | |
| "grad_norm": 0.24566735026281666, | |
| "learning_rate": 4.658378361122936e-05, | |
| "loss": 0.4991, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4987814784727864, | |
| "grad_norm": 0.18986094039099186, | |
| "learning_rate": 4.622108456977773e-05, | |
| "loss": 0.5029, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.511779041429732, | |
| "grad_norm": 0.23326957776437438, | |
| "learning_rate": 4.585786058721687e-05, | |
| "loss": 0.4995, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.5247766043866777, | |
| "grad_norm": 0.3107226298797518, | |
| "learning_rate": 4.549414231273633e-05, | |
| "loss": 0.5026, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.537774167343623, | |
| "grad_norm": 0.22719154064847447, | |
| "learning_rate": 4.512996043723453e-05, | |
| "loss": 0.5016, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5507717303005686, | |
| "grad_norm": 0.1833290747151392, | |
| "learning_rate": 4.476534569072895e-05, | |
| "loss": 0.5015, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.563769293257514, | |
| "grad_norm": 0.18170499366430282, | |
| "learning_rate": 4.440032883976318e-05, | |
| "loss": 0.4987, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.57676685621446, | |
| "grad_norm": 0.22937959483532433, | |
| "learning_rate": 4.403494068481074e-05, | |
| "loss": 0.497, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5897644191714053, | |
| "grad_norm": 0.20103120503973224, | |
| "learning_rate": 4.3669212057676145e-05, | |
| "loss": 0.5039, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.602761982128351, | |
| "grad_norm": 0.19185288266242695, | |
| "learning_rate": 4.33031738188933e-05, | |
| "loss": 0.494, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.6157595450852966, | |
| "grad_norm": 0.17010477360475862, | |
| "learning_rate": 4.293685685512142e-05, | |
| "loss": 0.4983, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.628757108042242, | |
| "grad_norm": 0.22826164302714383, | |
| "learning_rate": 4.257029207653881e-05, | |
| "loss": 0.4973, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.6417546709991875, | |
| "grad_norm": 0.1999552754693765, | |
| "learning_rate": 4.220351041423462e-05, | |
| "loss": 0.4965, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6547522339561334, | |
| "grad_norm": 0.16313332181232204, | |
| "learning_rate": 4.183654281759888e-05, | |
| "loss": 0.5064, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.667749796913079, | |
| "grad_norm": 0.2067882657675044, | |
| "learning_rate": 4.1469420251710905e-05, | |
| "loss": 0.5006, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.680747359870024, | |
| "grad_norm": 0.19027529910668617, | |
| "learning_rate": 4.110217369472649e-05, | |
| "loss": 0.505, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.69374492282697, | |
| "grad_norm": 0.2032049934064908, | |
| "learning_rate": 4.07348341352639e-05, | |
| "loss": 0.4931, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.7067424857839155, | |
| "grad_norm": 0.16941618101686762, | |
| "learning_rate": 4.0367432569789065e-05, | |
| "loss": 0.4932, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.719740048740861, | |
| "grad_norm": 0.17079677319469672, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4903, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.732737611697807, | |
| "grad_norm": 0.19810836153539363, | |
| "learning_rate": 3.963256743021095e-05, | |
| "loss": 0.4938, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7457351746547523, | |
| "grad_norm": 0.16838352156942718, | |
| "learning_rate": 3.92651658647361e-05, | |
| "loss": 0.4932, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7587327376116977, | |
| "grad_norm": 0.19393287722628685, | |
| "learning_rate": 3.889782630527353e-05, | |
| "loss": 0.501, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7717303005686436, | |
| "grad_norm": 0.2322999523417372, | |
| "learning_rate": 3.853057974828911e-05, | |
| "loss": 0.5028, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.784727863525589, | |
| "grad_norm": 0.17963554890231898, | |
| "learning_rate": 3.816345718240113e-05, | |
| "loss": 0.4966, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.7977254264825344, | |
| "grad_norm": 0.17972346659387642, | |
| "learning_rate": 3.779648958576538e-05, | |
| "loss": 0.4971, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.81072298943948, | |
| "grad_norm": 0.18567150272224223, | |
| "learning_rate": 3.74297079234612e-05, | |
| "loss": 0.5, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.8237205523964257, | |
| "grad_norm": 0.13962164511716743, | |
| "learning_rate": 3.706314314487859e-05, | |
| "loss": 0.5029, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.836718115353371, | |
| "grad_norm": 0.15495008282651787, | |
| "learning_rate": 3.669682618110671e-05, | |
| "loss": 0.4963, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.849715678310317, | |
| "grad_norm": 0.14169366024663888, | |
| "learning_rate": 3.6330787942323855e-05, | |
| "loss": 0.4975, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.8627132412672625, | |
| "grad_norm": 0.17858977296408154, | |
| "learning_rate": 3.5965059315189274e-05, | |
| "loss": 0.5001, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.875710804224208, | |
| "grad_norm": 0.16013806431323374, | |
| "learning_rate": 3.559967116023683e-05, | |
| "loss": 0.5015, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8887083671811533, | |
| "grad_norm": 0.17965163443416698, | |
| "learning_rate": 3.523465430927106e-05, | |
| "loss": 0.4978, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.901705930138099, | |
| "grad_norm": 0.180187277461439, | |
| "learning_rate": 3.4870039562765475e-05, | |
| "loss": 0.4974, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.9147034930950446, | |
| "grad_norm": 0.1838861258279284, | |
| "learning_rate": 3.4505857687263675e-05, | |
| "loss": 0.4921, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.9277010560519905, | |
| "grad_norm": 0.18286519398138992, | |
| "learning_rate": 3.414213941278314e-05, | |
| "loss": 0.5012, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.940698619008936, | |
| "grad_norm": 0.1574518878317098, | |
| "learning_rate": 3.377891543022229e-05, | |
| "loss": 0.4967, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9536961819658814, | |
| "grad_norm": 0.18504753095964077, | |
| "learning_rate": 3.341621638877064e-05, | |
| "loss": 0.5026, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.966693744922827, | |
| "grad_norm": 0.1576894962337943, | |
| "learning_rate": 3.305407289332279e-05, | |
| "loss": 0.4944, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9796913078797727, | |
| "grad_norm": 0.1467985899864142, | |
| "learning_rate": 3.269251550189573e-05, | |
| "loss": 0.4954, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.992688870836718, | |
| "grad_norm": 0.1515755081778688, | |
| "learning_rate": 3.2331574723050474e-05, | |
| "loss": 0.5042, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.0389926888708367, | |
| "grad_norm": 0.2765548168210541, | |
| "learning_rate": 3.197128101331764e-05, | |
| "loss": 0.4549, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 3.051990251827782, | |
| "grad_norm": 0.19466505214102667, | |
| "learning_rate": 3.161166477462759e-05, | |
| "loss": 0.4548, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 3.064987814784728, | |
| "grad_norm": 0.2880161957309376, | |
| "learning_rate": 3.125275635174497e-05, | |
| "loss": 0.4582, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 3.0779853777416735, | |
| "grad_norm": 0.23822280817514718, | |
| "learning_rate": 3.089458602970828e-05, | |
| "loss": 0.453, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.090982940698619, | |
| "grad_norm": 0.25800133083185917, | |
| "learning_rate": 3.0537184031274306e-05, | |
| "loss": 0.4547, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.1039805036555648, | |
| "grad_norm": 0.2384112705824712, | |
| "learning_rate": 3.0180580514368037e-05, | |
| "loss": 0.4586, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.11697806661251, | |
| "grad_norm": 0.24369875451596973, | |
| "learning_rate": 2.9824805569537747e-05, | |
| "loss": 0.4551, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.1299756295694556, | |
| "grad_norm": 0.24117461954872021, | |
| "learning_rate": 2.9469889217416045e-05, | |
| "loss": 0.4526, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.1429731925264015, | |
| "grad_norm": 0.20788905500839922, | |
| "learning_rate": 2.9115861406186593e-05, | |
| "loss": 0.4543, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.155970755483347, | |
| "grad_norm": 0.20899161474751476, | |
| "learning_rate": 2.8762752009057232e-05, | |
| "loss": 0.4523, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.1689683184402924, | |
| "grad_norm": 0.21136922172757902, | |
| "learning_rate": 2.841059082173902e-05, | |
| "loss": 0.4551, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.181965881397238, | |
| "grad_norm": 0.20064139424506391, | |
| "learning_rate": 2.805940755993223e-05, | |
| "loss": 0.4522, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1949634443541837, | |
| "grad_norm": 0.18215269955576766, | |
| "learning_rate": 2.770923185681878e-05, | |
| "loss": 0.4543, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.207961007311129, | |
| "grad_norm": 0.20092168632579746, | |
| "learning_rate": 2.7360093260561904e-05, | |
| "loss": 0.4547, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.2209585702680745, | |
| "grad_norm": 0.15760927431037322, | |
| "learning_rate": 2.7012021231812666e-05, | |
| "loss": 0.4485, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.2339561332250204, | |
| "grad_norm": 0.1905855267036173, | |
| "learning_rate": 2.6665045141224193e-05, | |
| "loss": 0.4514, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.246953696181966, | |
| "grad_norm": 0.15059676552502058, | |
| "learning_rate": 2.6319194266973256e-05, | |
| "loss": 0.4587, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.2599512591389113, | |
| "grad_norm": 0.18536963620839997, | |
| "learning_rate": 2.597449779228983e-05, | |
| "loss": 0.4534, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.272948822095857, | |
| "grad_norm": 0.17274276141163067, | |
| "learning_rate": 2.563098480299451e-05, | |
| "loss": 0.457, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2859463850528026, | |
| "grad_norm": 0.1562543243595631, | |
| "learning_rate": 2.5288684285044283e-05, | |
| "loss": 0.4549, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.298943948009748, | |
| "grad_norm": 0.17358175772296236, | |
| "learning_rate": 2.4947625122086585e-05, | |
| "loss": 0.4548, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.311941510966694, | |
| "grad_norm": 0.17184079642105152, | |
| "learning_rate": 2.460783609302218e-05, | |
| "loss": 0.4521, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.3249390739236393, | |
| "grad_norm": 0.16030118314009942, | |
| "learning_rate": 2.4269345869576676e-05, | |
| "loss": 0.4469, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.3379366368805847, | |
| "grad_norm": 0.16681227025456086, | |
| "learning_rate": 2.393218301388123e-05, | |
| "loss": 0.4532, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.3509341998375306, | |
| "grad_norm": 0.13889802201519397, | |
| "learning_rate": 2.35963759760624e-05, | |
| "loss": 0.4524, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.363931762794476, | |
| "grad_norm": 0.15349090525146497, | |
| "learning_rate": 2.3261953091841553e-05, | |
| "loss": 0.4546, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.3769293257514215, | |
| "grad_norm": 0.13644110770797446, | |
| "learning_rate": 2.2928942580143855e-05, | |
| "loss": 0.4526, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3899268887083673, | |
| "grad_norm": 0.1490492694969517, | |
| "learning_rate": 2.2597372540717083e-05, | |
| "loss": 0.4568, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.4029244516653128, | |
| "grad_norm": 0.14894023043131757, | |
| "learning_rate": 2.226727095176057e-05, | |
| "loss": 0.4588, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.415922014622258, | |
| "grad_norm": 0.14995590721137517, | |
| "learning_rate": 2.1938665667564435e-05, | |
| "loss": 0.4507, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.428919577579204, | |
| "grad_norm": 0.14184790166599112, | |
| "learning_rate": 2.1611584416159106e-05, | |
| "loss": 0.4496, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.4419171405361495, | |
| "grad_norm": 0.13481813074778676, | |
| "learning_rate": 2.1286054796975696e-05, | |
| "loss": 0.4541, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.454914703493095, | |
| "grad_norm": 0.12999426318312995, | |
| "learning_rate": 2.096210427851706e-05, | |
| "loss": 0.4544, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.467912266450041, | |
| "grad_norm": 0.13239806186780637, | |
| "learning_rate": 2.063976019604006e-05, | |
| "loss": 0.451, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.4809098294069862, | |
| "grad_norm": 0.1271255029835273, | |
| "learning_rate": 2.0319049749248876e-05, | |
| "loss": 0.4487, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4939073923639317, | |
| "grad_norm": 0.13303879175543112, | |
| "learning_rate": 2.0000000000000012e-05, | |
| "loss": 0.4522, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.506904955320877, | |
| "grad_norm": 0.12772791682528986, | |
| "learning_rate": 1.9682637870018638e-05, | |
| "loss": 0.4486, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.519902518277823, | |
| "grad_norm": 0.13368037803272537, | |
| "learning_rate": 1.9366990138627054e-05, | |
| "loss": 0.4556, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.5329000812347684, | |
| "grad_norm": 0.12998443969153295, | |
| "learning_rate": 1.9053083440484887e-05, | |
| "loss": 0.4573, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.5458976441917143, | |
| "grad_norm": 0.1299143292827015, | |
| "learning_rate": 1.8740944263341773e-05, | |
| "loss": 0.4496, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.5588952071486597, | |
| "grad_norm": 0.13501723708928842, | |
| "learning_rate": 1.8430598945802156e-05, | |
| "loss": 0.4529, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.571892770105605, | |
| "grad_norm": 0.1392141581263843, | |
| "learning_rate": 1.8122073675102935e-05, | |
| "loss": 0.4606, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.5848903330625506, | |
| "grad_norm": 0.13279583183929072, | |
| "learning_rate": 1.781539448490365e-05, | |
| "loss": 0.4559, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5978878960194964, | |
| "grad_norm": 0.13802766481193707, | |
| "learning_rate": 1.7510587253089842e-05, | |
| "loss": 0.4525, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.610885458976442, | |
| "grad_norm": 0.1335727117344679, | |
| "learning_rate": 1.7207677699589355e-05, | |
| "loss": 0.4529, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.6238830219333877, | |
| "grad_norm": 0.12489253393946519, | |
| "learning_rate": 1.690669138420215e-05, | |
| "loss": 0.4582, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.636880584890333, | |
| "grad_norm": 0.1413569401198253, | |
| "learning_rate": 1.6607653704443457e-05, | |
| "loss": 0.4559, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.6498781478472786, | |
| "grad_norm": 0.11828655169190146, | |
| "learning_rate": 1.6310589893400804e-05, | |
| "loss": 0.4518, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.662875710804224, | |
| "grad_norm": 0.12039871469350676, | |
| "learning_rate": 1.601552501760473e-05, | |
| "loss": 0.4513, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.67587327376117, | |
| "grad_norm": 0.10972016559298864, | |
| "learning_rate": 1.5722483974913737e-05, | |
| "loss": 0.4585, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.6888708367181153, | |
| "grad_norm": 0.11605815211359002, | |
| "learning_rate": 1.5431491492413288e-05, | |
| "loss": 0.4548, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.7018683996750608, | |
| "grad_norm": 0.11113738686213914, | |
| "learning_rate": 1.5142572124329418e-05, | |
| "loss": 0.456, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.7148659626320066, | |
| "grad_norm": 0.11801948433211017, | |
| "learning_rate": 1.4855750249956718e-05, | |
| "loss": 0.4567, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.727863525588952, | |
| "grad_norm": 0.11508267709388358, | |
| "learning_rate": 1.457105007160129e-05, | |
| "loss": 0.4567, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.7408610885458975, | |
| "grad_norm": 0.11951109036290335, | |
| "learning_rate": 1.4288495612538427e-05, | |
| "loss": 0.4543, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.7538586515028434, | |
| "grad_norm": 0.11609444931011924, | |
| "learning_rate": 1.4008110714985623e-05, | |
| "loss": 0.457, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.766856214459789, | |
| "grad_norm": 0.12161842365979787, | |
| "learning_rate": 1.3729919038090627e-05, | |
| "loss": 0.4574, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.7798537774167342, | |
| "grad_norm": 0.12517775168843562, | |
| "learning_rate": 1.3453944055935151e-05, | |
| "loss": 0.4547, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.79285134037368, | |
| "grad_norm": 0.11520911340042557, | |
| "learning_rate": 1.3180209055554043e-05, | |
| "loss": 0.4513, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.8058489033306255, | |
| "grad_norm": 0.11815010738675953, | |
| "learning_rate": 1.2908737134970367e-05, | |
| "loss": 0.443, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.818846466287571, | |
| "grad_norm": 0.1162507897952791, | |
| "learning_rate": 1.2639551201246278e-05, | |
| "loss": 0.4516, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.8318440292445164, | |
| "grad_norm": 0.11366641746047221, | |
| "learning_rate": 1.2372673968550229e-05, | |
| "loss": 0.4547, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.8448415922014623, | |
| "grad_norm": 0.13504962259021788, | |
| "learning_rate": 1.2108127956240186e-05, | |
| "loss": 0.4455, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.8578391551584077, | |
| "grad_norm": 0.11026952373517614, | |
| "learning_rate": 1.1845935486963546e-05, | |
| "loss": 0.4525, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.8708367181153536, | |
| "grad_norm": 0.10668225682868658, | |
| "learning_rate": 1.158611868477344e-05, | |
| "loss": 0.4566, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.883834281072299, | |
| "grad_norm": 0.1354540165594788, | |
| "learning_rate": 1.1328699473261957e-05, | |
| "loss": 0.4472, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8968318440292444, | |
| "grad_norm": 0.10921111595164792, | |
| "learning_rate": 1.107369957371013e-05, | |
| "loss": 0.4473, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.90982940698619, | |
| "grad_norm": 0.10437459191229037, | |
| "learning_rate": 1.0821140503255174e-05, | |
| "loss": 0.4539, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.9228269699431357, | |
| "grad_norm": 0.11124834426391973, | |
| "learning_rate": 1.0571043573074737e-05, | |
| "loss": 0.4524, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.935824532900081, | |
| "grad_norm": 0.11103073467741553, | |
| "learning_rate": 1.0323429886588743e-05, | |
| "loss": 0.4551, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.948822095857027, | |
| "grad_norm": 0.0996775517579157, | |
| "learning_rate": 1.0078320337678584e-05, | |
| "loss": 0.4542, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.9618196588139725, | |
| "grad_norm": 0.10553906139080531, | |
| "learning_rate": 9.835735608924155e-06, | |
| "loss": 0.4559, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.974817221770918, | |
| "grad_norm": 0.11276985701526075, | |
| "learning_rate": 9.595696169858542e-06, | |
| "loss": 0.4556, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.9878147847278633, | |
| "grad_norm": 0.10020067885290669, | |
| "learning_rate": 9.358222275240884e-06, | |
| "loss": 0.452, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 4.000812347684809, | |
| "grad_norm": 0.1015888198294778, | |
| "learning_rate": 9.123333963347166e-06, | |
| "loss": 0.4523, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 4.013809910641755, | |
| "grad_norm": 0.19346874584833063, | |
| "learning_rate": 8.89105105427945e-06, | |
| "loss": 0.4304, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 4.0268074735987005, | |
| "grad_norm": 0.1492042453109882, | |
| "learning_rate": 8.661393148293355e-06, | |
| "loss": 0.4184, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 4.039805036555646, | |
| "grad_norm": 0.11128451836895309, | |
| "learning_rate": 8.434379624144261e-06, | |
| "loss": 0.4272, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 4.052802599512591, | |
| "grad_norm": 0.15566082975695297, | |
| "learning_rate": 8.210029637452016e-06, | |
| "loss": 0.4291, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 4.065800162469537, | |
| "grad_norm": 0.18015489370960358, | |
| "learning_rate": 7.988362119084642e-06, | |
| "loss": 0.43, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 4.078797725426482, | |
| "grad_norm": 0.16617077404145308, | |
| "learning_rate": 7.769395773560874e-06, | |
| "loss": 0.4245, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 4.0917952883834285, | |
| "grad_norm": 0.12915804996415034, | |
| "learning_rate": 7.553149077471915e-06, | |
| "loss": 0.4205, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.104792851340374, | |
| "grad_norm": 0.12694294053103128, | |
| "learning_rate": 7.3396402779222845e-06, | |
| "loss": 0.4281, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.117790414297319, | |
| "grad_norm": 0.14590711745883797, | |
| "learning_rate": 7.128887390990198e-06, | |
| "loss": 0.4361, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.130787977254265, | |
| "grad_norm": 0.13808851249499354, | |
| "learning_rate": 6.9209082002072725e-06, | |
| "loss": 0.4215, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.14378554021121, | |
| "grad_norm": 0.129695250623056, | |
| "learning_rate": 6.715720255058e-06, | |
| "loss": 0.4252, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.156783103168156, | |
| "grad_norm": 0.11617214795717144, | |
| "learning_rate": 6.513340869498859e-06, | |
| "loss": 0.4251, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.169780666125101, | |
| "grad_norm": 0.12511546150804922, | |
| "learning_rate": 6.313787120497376e-06, | |
| "loss": 0.4223, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.182778229082047, | |
| "grad_norm": 0.11917710079150087, | |
| "learning_rate": 6.117075846591123e-06, | |
| "loss": 0.4259, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.195775792038993, | |
| "grad_norm": 0.11982991360751519, | |
| "learning_rate": 5.923223646466923e-06, | |
| "loss": 0.4267, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.208773354995938, | |
| "grad_norm": 0.11087809076543045, | |
| "learning_rate": 5.732246877560146e-06, | |
| "loss": 0.4237, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.221770917952884, | |
| "grad_norm": 0.11037591637224678, | |
| "learning_rate": 5.5441616546745646e-06, | |
| "loss": 0.4253, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.234768480909829, | |
| "grad_norm": 0.11290723637234998, | |
| "learning_rate": 5.358983848622452e-06, | |
| "loss": 0.4272, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.247766043866775, | |
| "grad_norm": 0.11769455319020229, | |
| "learning_rate": 5.176729084885508e-06, | |
| "loss": 0.4234, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.260763606823721, | |
| "grad_norm": 0.11385474199844745, | |
| "learning_rate": 4.99741274229625e-06, | |
| "loss": 0.4323, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.273761169780666, | |
| "grad_norm": 0.10749198484614475, | |
| "learning_rate": 4.821049951740442e-06, | |
| "loss": 0.4295, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.286758732737612, | |
| "grad_norm": 0.10566297009015177, | |
| "learning_rate": 4.647655594880225e-06, | |
| "loss": 0.4254, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.299756295694557, | |
| "grad_norm": 0.10419068782938447, | |
| "learning_rate": 4.4772443028985004e-06, | |
| "loss": 0.4252, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.312753858651503, | |
| "grad_norm": 0.10950077630402558, | |
| "learning_rate": 4.3098304552642385e-06, | |
| "loss": 0.4242, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.325751421608448, | |
| "grad_norm": 0.1100371440360915, | |
| "learning_rate": 4.1454281785191995e-06, | |
| "loss": 0.4197, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.338748984565394, | |
| "grad_norm": 0.10391792774677544, | |
| "learning_rate": 3.984051345085855e-06, | |
| "loss": 0.4258, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.35174654752234, | |
| "grad_norm": 0.09904939283103152, | |
| "learning_rate": 3.825713572096903e-06, | |
| "loss": 0.4184, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.364744110479285, | |
| "grad_norm": 0.10192004321049586, | |
| "learning_rate": 3.6704282202461515e-06, | |
| "loss": 0.4281, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.377741673436231, | |
| "grad_norm": 0.09890117596157007, | |
| "learning_rate": 3.518208392661184e-06, | |
| "loss": 0.4197, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.390739236393176, | |
| "grad_norm": 0.09792268745311886, | |
| "learning_rate": 3.3690669337977e-06, | |
| "loss": 0.4265, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.4037367993501215, | |
| "grad_norm": 0.09563298489182019, | |
| "learning_rate": 3.2230164283556918e-06, | |
| "loss": 0.4251, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.416734362307068, | |
| "grad_norm": 0.09937136144673961, | |
| "learning_rate": 3.080069200217497e-06, | |
| "loss": 0.4345, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.429731925264013, | |
| "grad_norm": 0.09928349883535201, | |
| "learning_rate": 2.9402373114079295e-06, | |
| "loss": 0.4285, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.442729488220959, | |
| "grad_norm": 0.09247546072284324, | |
| "learning_rate": 2.803532561076492e-06, | |
| "loss": 0.4273, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.455727051177904, | |
| "grad_norm": 0.09749820385703685, | |
| "learning_rate": 2.669966484501716e-06, | |
| "loss": 0.4269, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.46872461413485, | |
| "grad_norm": 0.09646403701008398, | |
| "learning_rate": 2.5395503521178143e-06, | |
| "loss": 0.4308, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.481722177091795, | |
| "grad_norm": 0.0936333573004958, | |
| "learning_rate": 2.4122951685636674e-06, | |
| "loss": 0.4224, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.49471974004874, | |
| "grad_norm": 0.09222486151634639, | |
| "learning_rate": 2.2882116717542634e-06, | |
| "loss": 0.4228, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.507717303005687, | |
| "grad_norm": 0.0937700330761675, | |
| "learning_rate": 2.1673103319746146e-06, | |
| "loss": 0.4251, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.520714865962632, | |
| "grad_norm": 0.09523262003402952, | |
| "learning_rate": 2.049601350996233e-06, | |
| "loss": 0.425, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.533712428919578, | |
| "grad_norm": 0.09580115361604435, | |
| "learning_rate": 1.93509466121633e-06, | |
| "loss": 0.4278, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.546709991876523, | |
| "grad_norm": 0.09551827606861985, | |
| "learning_rate": 1.8237999248197002e-06, | |
| "loss": 0.4277, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.5597075548334685, | |
| "grad_norm": 0.09398206095795404, | |
| "learning_rate": 1.7157265329634354e-06, | |
| "loss": 0.4296, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.572705117790415, | |
| "grad_norm": 0.09186446471518088, | |
| "learning_rate": 1.6108836049844434e-06, | |
| "loss": 0.4254, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.58570268074736, | |
| "grad_norm": 0.09291207003002876, | |
| "learning_rate": 1.5092799876299835e-06, | |
| "loss": 0.4228, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.598700243704306, | |
| "grad_norm": 0.09321080065504848, | |
| "learning_rate": 1.4109242543111834e-06, | |
| "loss": 0.4224, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.611697806661251, | |
| "grad_norm": 0.09820369207399382, | |
| "learning_rate": 1.3158247043795735e-06, | |
| "loss": 0.425, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.6246953696181965, | |
| "grad_norm": 0.09483780290747863, | |
| "learning_rate": 1.2239893624267852e-06, | |
| "loss": 0.4248, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.637692932575142, | |
| "grad_norm": 0.09731358909691826, | |
| "learning_rate": 1.1354259776074472e-06, | |
| "loss": 0.4235, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.650690495532087, | |
| "grad_norm": 0.0904965273633941, | |
| "learning_rate": 1.050142022985292e-06, | |
| "loss": 0.4181, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.663688058489034, | |
| "grad_norm": 0.08967404436562013, | |
| "learning_rate": 9.681446949025752e-07, | |
| "loss": 0.4294, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.676685621445979, | |
| "grad_norm": 0.09202959723546777, | |
| "learning_rate": 8.89440912372832e-07, | |
| "loss": 0.4223, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.6896831844029245, | |
| "grad_norm": 0.0934294923741134, | |
| "learning_rate": 8.140373164970428e-07, | |
| "loss": 0.4254, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.70268074735987, | |
| "grad_norm": 0.09277475785552658, | |
| "learning_rate": 7.419402699032852e-07, | |
| "loss": 0.4264, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.715678310316815, | |
| "grad_norm": 0.0963901158463573, | |
| "learning_rate": 6.731558562097995e-07, | |
| "loss": 0.4244, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.728675873273761, | |
| "grad_norm": 0.08814663638316801, | |
| "learning_rate": 6.076898795116792e-07, | |
| "loss": 0.4191, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.741673436230707, | |
| "grad_norm": 0.08994523311851658, | |
| "learning_rate": 5.455478638911071e-07, | |
| "loss": 0.4287, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.754670999187653, | |
| "grad_norm": 0.08992860659969752, | |
| "learning_rate": 4.867350529512261e-07, | |
| "loss": 0.4243, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.767668562144598, | |
| "grad_norm": 0.08903822534894691, | |
| "learning_rate": 4.3125640937368373e-07, | |
| "loss": 0.4254, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.780666125101543, | |
| "grad_norm": 0.09037872867259723, | |
| "learning_rate": 3.791166144998704e-07, | |
| "loss": 0.4237, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.793663688058489, | |
| "grad_norm": 0.08849981605290622, | |
| "learning_rate": 3.3032006793590977e-07, | |
| "loss": 0.4256, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.806661251015434, | |
| "grad_norm": 0.08939190625410524, | |
| "learning_rate": 2.848708871814054e-07, | |
| "loss": 0.432, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.81965881397238, | |
| "grad_norm": 0.08978512061239115, | |
| "learning_rate": 2.4277290728202063e-07, | |
| "loss": 0.431, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.832656376929326, | |
| "grad_norm": 0.08749342151530978, | |
| "learning_rate": 2.040296805058528e-07, | |
| "loss": 0.4282, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.8456539398862715, | |
| "grad_norm": 0.09394237513340908, | |
| "learning_rate": 1.6864447604370004e-07, | |
| "loss": 0.4175, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.858651502843217, | |
| "grad_norm": 0.0881034945018527, | |
| "learning_rate": 1.3662027973320614e-07, | |
| "loss": 0.4225, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.871649065800162, | |
| "grad_norm": 0.08852561609870382, | |
| "learning_rate": 1.0795979380690657e-07, | |
| "loss": 0.4227, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.884646628757108, | |
| "grad_norm": 0.0911620154787333, | |
| "learning_rate": 8.266543666421544e-08, | |
| "loss": 0.429, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.897644191714054, | |
| "grad_norm": 0.08685734979527214, | |
| "learning_rate": 6.073934266735303e-08, | |
| "loss": 0.4251, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.9106417546709995, | |
| "grad_norm": 0.08736267011789721, | |
| "learning_rate": 4.218336196125439e-08, | |
| "loss": 0.4222, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.923639317627945, | |
| "grad_norm": 0.0888955083475754, | |
| "learning_rate": 2.699906031745414e-08, | |
| "loss": 0.4263, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.93663688058489, | |
| "grad_norm": 0.08888998189271619, | |
| "learning_rate": 1.5187719001943378e-08, | |
| "loss": 0.429, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.949634443541836, | |
| "grad_norm": 0.08806089423257202, | |
| "learning_rate": 6.750334667091629e-09, | |
| "loss": 0.4241, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.962632006498781, | |
| "grad_norm": 0.0864294126107438, | |
| "learning_rate": 1.6876192675052695e-09, | |
| "loss": 0.4205, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.975629569455727, | |
| "grad_norm": 0.0892424352366353, | |
| "learning_rate": 0.0, | |
| "loss": 0.4228, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.975629569455727, | |
| "step": 380, | |
| "total_flos": 9.790538054727893e+18, | |
| "train_loss": 0.17350224256515503, | |
| "train_runtime": 16013.9297, | |
| "train_samples_per_second": 12.293, | |
| "train_steps_per_second": 0.024 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 380, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 9.790538054727893e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |