| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 4.94557270511779, |
| "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": 6.099783948625711, |
| "learning_rate": 2.105263157894737e-06, |
| "loss": 1.0219, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.025995125913891144, |
| "grad_norm": 6.068915599090855, |
| "learning_rate": 4.210526315789474e-06, |
| "loss": 1.0148, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.03899268887083672, |
| "grad_norm": 5.594712813142529, |
| "learning_rate": 6.31578947368421e-06, |
| "loss": 0.9975, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.05199025182778229, |
| "grad_norm": 4.038714211131839, |
| "learning_rate": 8.421052631578948e-06, |
| "loss": 0.9696, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.06498781478472786, |
| "grad_norm": 2.365897221739178, |
| "learning_rate": 1.0526315789473684e-05, |
| "loss": 0.9322, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.07798537774167344, |
| "grad_norm": 4.808874284957715, |
| "learning_rate": 1.263157894736842e-05, |
| "loss": 0.9538, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.09098294069861901, |
| "grad_norm": 4.840119194557309, |
| "learning_rate": 1.4736842105263159e-05, |
| "loss": 0.9456, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.10398050365556458, |
| "grad_norm": 5.7169851588504645, |
| "learning_rate": 1.6842105263157896e-05, |
| "loss": 0.9349, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.11697806661251016, |
| "grad_norm": 4.120707061579036, |
| "learning_rate": 1.894736842105263e-05, |
| "loss": 0.9107, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.12997562956945571, |
| "grad_norm": 2.7497360309830476, |
| "learning_rate": 2.105263157894737e-05, |
| "loss": 0.8669, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.1429731925264013, |
| "grad_norm": 1.8093148555009293, |
| "learning_rate": 2.3157894736842107e-05, |
| "loss": 0.8289, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.15597075548334688, |
| "grad_norm": 1.617836052019221, |
| "learning_rate": 2.526315789473684e-05, |
| "loss": 0.8096, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.16896831844029245, |
| "grad_norm": 1.0744851220126015, |
| "learning_rate": 2.7368421052631583e-05, |
| "loss": 0.7877, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.18196588139723802, |
| "grad_norm": 1.2705200346488472, |
| "learning_rate": 2.9473684210526317e-05, |
| "loss": 0.78, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.19496344435418358, |
| "grad_norm": 1.4130680513032683, |
| "learning_rate": 3.157894736842106e-05, |
| "loss": 0.7769, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.20796100731112915, |
| "grad_norm": 1.324813747526517, |
| "learning_rate": 3.368421052631579e-05, |
| "loss": 0.7705, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.22095857026807472, |
| "grad_norm": 1.2590819742707091, |
| "learning_rate": 3.578947368421053e-05, |
| "loss": 0.7458, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.23395613322502032, |
| "grad_norm": 1.2135489907048478, |
| "learning_rate": 3.789473684210526e-05, |
| "loss": 0.7415, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.2469536961819659, |
| "grad_norm": 0.955476297443492, |
| "learning_rate": 4e-05, |
| "loss": 0.7311, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.25995125913891143, |
| "grad_norm": 1.6971617462330089, |
| "learning_rate": 4.210526315789474e-05, |
| "loss": 0.7413, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.272948822095857, |
| "grad_norm": 1.4395879957659508, |
| "learning_rate": 4.421052631578948e-05, |
| "loss": 0.7295, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.2859463850528026, |
| "grad_norm": 1.171007731697555, |
| "learning_rate": 4.6315789473684214e-05, |
| "loss": 0.7153, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.29894394800974816, |
| "grad_norm": 1.931104084566943, |
| "learning_rate": 4.842105263157895e-05, |
| "loss": 0.7249, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.31194151096669376, |
| "grad_norm": 1.1062671298809976, |
| "learning_rate": 5.052631578947368e-05, |
| "loss": 0.7178, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.3249390739236393, |
| "grad_norm": 2.0237024200236986, |
| "learning_rate": 5.263157894736843e-05, |
| "loss": 0.7214, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.3379366368805849, |
| "grad_norm": 1.6595941751589156, |
| "learning_rate": 5.4736842105263165e-05, |
| "loss": 0.702, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.35093419983753044, |
| "grad_norm": 1.643241943721994, |
| "learning_rate": 5.68421052631579e-05, |
| "loss": 0.7054, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.36393176279447603, |
| "grad_norm": 1.9552851038948404, |
| "learning_rate": 5.8947368421052634e-05, |
| "loss": 0.7013, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.37692932575142163, |
| "grad_norm": 1.1207858673487592, |
| "learning_rate": 6.105263157894738e-05, |
| "loss": 0.6901, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.38992688870836717, |
| "grad_norm": 2.1963019920042757, |
| "learning_rate": 6.315789473684212e-05, |
| "loss": 0.6868, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.40292445166531277, |
| "grad_norm": 1.4781504300412824, |
| "learning_rate": 6.526315789473685e-05, |
| "loss": 0.6903, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.4159220146222583, |
| "grad_norm": 2.122635652457263, |
| "learning_rate": 6.736842105263159e-05, |
| "loss": 0.6858, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.4289195775792039, |
| "grad_norm": 1.650492771547578, |
| "learning_rate": 6.947368421052632e-05, |
| "loss": 0.6898, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.44191714053614944, |
| "grad_norm": 1.5236371764651064, |
| "learning_rate": 7.157894736842105e-05, |
| "loss": 0.6813, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.45491470349309504, |
| "grad_norm": 1.185803755084194, |
| "learning_rate": 7.368421052631579e-05, |
| "loss": 0.6829, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.46791226645004064, |
| "grad_norm": 1.9405604072257883, |
| "learning_rate": 7.578947368421052e-05, |
| "loss": 0.6738, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.4809098294069862, |
| "grad_norm": 1.363411287783038, |
| "learning_rate": 7.789473684210527e-05, |
| "loss": 0.6696, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.4939073923639318, |
| "grad_norm": 2.0874162228533115, |
| "learning_rate": 8e-05, |
| "loss": 0.6753, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.5069049553208773, |
| "grad_norm": 1.8491550544598394, |
| "learning_rate": 7.99983123807325e-05, |
| "loss": 0.6723, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.5199025182778229, |
| "grad_norm": 1.2473648539637896, |
| "learning_rate": 7.999324966533291e-05, |
| "loss": 0.66, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.5329000812347685, |
| "grad_norm": 3.3198725973160084, |
| "learning_rate": 7.998481228099806e-05, |
| "loss": 0.675, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.545897644191714, |
| "grad_norm": 2.604690756502679, |
| "learning_rate": 7.997300093968255e-05, |
| "loss": 0.6824, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.5588952071486596, |
| "grad_norm": 2.02050667371959, |
| "learning_rate": 7.995781663803876e-05, |
| "loss": 0.6667, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.5718927701056052, |
| "grad_norm": 1.1560247057038098, |
| "learning_rate": 7.993926065733265e-05, |
| "loss": 0.6619, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.5848903330625508, |
| "grad_norm": 1.5367530120814719, |
| "learning_rate": 7.991733456333579e-05, |
| "loss": 0.6664, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.5978878960194963, |
| "grad_norm": 1.1634220605553418, |
| "learning_rate": 7.98920402061931e-05, |
| "loss": 0.6493, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.6108854589764419, |
| "grad_norm": 1.015353240524831, |
| "learning_rate": 7.98633797202668e-05, |
| "loss": 0.659, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.6238830219333875, |
| "grad_norm": 1.4952317667466968, |
| "learning_rate": 7.98313555239563e-05, |
| "loss": 0.6609, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.6368805848903331, |
| "grad_norm": 0.8908692378681831, |
| "learning_rate": 7.979597031949415e-05, |
| "loss": 0.6638, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.6498781478472786, |
| "grad_norm": 1.7121745994414552, |
| "learning_rate": 7.975722709271799e-05, |
| "loss": 0.6636, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.6628757108042242, |
| "grad_norm": 1.2827332360089716, |
| "learning_rate": 7.97151291128186e-05, |
| "loss": 0.6562, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.6758732737611698, |
| "grad_norm": 1.1190183435245546, |
| "learning_rate": 7.96696799320641e-05, |
| "loss": 0.6467, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.6888708367181153, |
| "grad_norm": 1.5796475425952419, |
| "learning_rate": 7.962088338550013e-05, |
| "loss": 0.6549, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.7018683996750609, |
| "grad_norm": 1.0356021868877745, |
| "learning_rate": 7.956874359062632e-05, |
| "loss": 0.6486, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.7148659626320065, |
| "grad_norm": 1.689157245211179, |
| "learning_rate": 7.951326494704878e-05, |
| "loss": 0.6427, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.7278635255889521, |
| "grad_norm": 1.3036787434531885, |
| "learning_rate": 7.94544521361089e-05, |
| "loss": 0.6451, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.7408610885458976, |
| "grad_norm": 1.411604156835551, |
| "learning_rate": 7.939231012048833e-05, |
| "loss": 0.6434, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.7538586515028433, |
| "grad_norm": 1.073329384486516, |
| "learning_rate": 7.932684414379021e-05, |
| "loss": 0.6417, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.7668562144597888, |
| "grad_norm": 1.1364966234135832, |
| "learning_rate": 7.925805973009672e-05, |
| "loss": 0.6348, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.7798537774167343, |
| "grad_norm": 0.9594628748904098, |
| "learning_rate": 7.918596268350296e-05, |
| "loss": 0.6399, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.7928513403736799, |
| "grad_norm": 1.3124733925076593, |
| "learning_rate": 7.911055908762718e-05, |
| "loss": 0.6445, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.8058489033306255, |
| "grad_norm": 0.8150161261530253, |
| "learning_rate": 7.903185530509743e-05, |
| "loss": 0.6333, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.8188464662875711, |
| "grad_norm": 0.8675788545202524, |
| "learning_rate": 7.894985797701472e-05, |
| "loss": 0.6342, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.8318440292445166, |
| "grad_norm": 0.8701303058033523, |
| "learning_rate": 7.886457402239256e-05, |
| "loss": 0.627, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.8448415922014623, |
| "grad_norm": 1.1211379390083465, |
| "learning_rate": 7.877601063757323e-05, |
| "loss": 0.6346, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.8578391551584078, |
| "grad_norm": 1.0237659842661562, |
| "learning_rate": 7.868417529562043e-05, |
| "loss": 0.6244, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.8708367181153533, |
| "grad_norm": 0.8724494077448317, |
| "learning_rate": 7.858907574568882e-05, |
| "loss": 0.6279, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.8838342810722989, |
| "grad_norm": 0.8890596583749399, |
| "learning_rate": 7.849072001237001e-05, |
| "loss": 0.6212, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.8968318440292445, |
| "grad_norm": 0.9491308234698868, |
| "learning_rate": 7.838911639501557e-05, |
| "loss": 0.6164, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.9098294069861901, |
| "grad_norm": 0.9144632388033332, |
| "learning_rate": 7.828427346703657e-05, |
| "loss": 0.6141, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.9228269699431356, |
| "grad_norm": 1.0369164699296847, |
| "learning_rate": 7.81762000751803e-05, |
| "loss": 0.6227, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.9358245329000813, |
| "grad_norm": 1.0824992143715033, |
| "learning_rate": 7.806490533878368e-05, |
| "loss": 0.6219, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.9488220958570268, |
| "grad_norm": 0.8841365960283825, |
| "learning_rate": 7.795039864900378e-05, |
| "loss": 0.6134, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.9618196588139724, |
| "grad_norm": 0.6911046945305666, |
| "learning_rate": 7.783268966802539e-05, |
| "loss": 0.6263, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.974817221770918, |
| "grad_norm": 0.7831186784164668, |
| "learning_rate": 7.771178832824573e-05, |
| "loss": 0.6133, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.9878147847278635, |
| "grad_norm": 0.6782592588604553, |
| "learning_rate": 7.758770483143634e-05, |
| "loss": 0.6102, |
| "step": 76 |
| }, |
| { |
| "epoch": 1.0024370430544274, |
| "grad_norm": 0.6336474343134532, |
| "learning_rate": 7.74604496478822e-05, |
| "loss": 0.6116, |
| "step": 77 |
| }, |
| { |
| "epoch": 1.0154346060113728, |
| "grad_norm": 0.6278277714102893, |
| "learning_rate": 7.733003351549829e-05, |
| "loss": 0.5828, |
| "step": 78 |
| }, |
| { |
| "epoch": 1.0284321689683185, |
| "grad_norm": 0.6209449077089193, |
| "learning_rate": 7.719646743892352e-05, |
| "loss": 0.584, |
| "step": 79 |
| }, |
| { |
| "epoch": 1.0414297319252641, |
| "grad_norm": 0.73979662934484, |
| "learning_rate": 7.705976268859207e-05, |
| "loss": 0.5892, |
| "step": 80 |
| }, |
| { |
| "epoch": 1.0544272948822095, |
| "grad_norm": 0.8782533057922014, |
| "learning_rate": 7.691993079978252e-05, |
| "loss": 0.5809, |
| "step": 81 |
| }, |
| { |
| "epoch": 1.0674248578391552, |
| "grad_norm": 0.9723345888414255, |
| "learning_rate": 7.677698357164431e-05, |
| "loss": 0.5745, |
| "step": 82 |
| }, |
| { |
| "epoch": 1.0804224207961006, |
| "grad_norm": 0.9060339939969984, |
| "learning_rate": 7.663093306620231e-05, |
| "loss": 0.5768, |
| "step": 83 |
| }, |
| { |
| "epoch": 1.0934199837530463, |
| "grad_norm": 0.6635148814822602, |
| "learning_rate": 7.648179160733883e-05, |
| "loss": 0.5735, |
| "step": 84 |
| }, |
| { |
| "epoch": 1.106417546709992, |
| "grad_norm": 0.626645192938145, |
| "learning_rate": 7.632957177975387e-05, |
| "loss": 0.5774, |
| "step": 85 |
| }, |
| { |
| "epoch": 1.1194151096669374, |
| "grad_norm": 0.5880549876089324, |
| "learning_rate": 7.61742864279031e-05, |
| "loss": 0.5806, |
| "step": 86 |
| }, |
| { |
| "epoch": 1.132412672623883, |
| "grad_norm": 0.7256826177123274, |
| "learning_rate": 7.601594865491414e-05, |
| "loss": 0.5725, |
| "step": 87 |
| }, |
| { |
| "epoch": 1.1454102355808287, |
| "grad_norm": 1.0226725048413887, |
| "learning_rate": 7.585457182148081e-05, |
| "loss": 0.5766, |
| "step": 88 |
| }, |
| { |
| "epoch": 1.158407798537774, |
| "grad_norm": 1.053827253244498, |
| "learning_rate": 7.569016954473577e-05, |
| "loss": 0.5781, |
| "step": 89 |
| }, |
| { |
| "epoch": 1.1714053614947197, |
| "grad_norm": 0.9313744365954294, |
| "learning_rate": 7.552275569710152e-05, |
| "loss": 0.5826, |
| "step": 90 |
| }, |
| { |
| "epoch": 1.1844029244516654, |
| "grad_norm": 1.054632648598746, |
| "learning_rate": 7.535234440511979e-05, |
| "loss": 0.581, |
| "step": 91 |
| }, |
| { |
| "epoch": 1.1974004874086108, |
| "grad_norm": 0.8836759025467023, |
| "learning_rate": 7.517895004825956e-05, |
| "loss": 0.5821, |
| "step": 92 |
| }, |
| { |
| "epoch": 1.2103980503655565, |
| "grad_norm": 1.127228102294561, |
| "learning_rate": 7.500258725770375e-05, |
| "loss": 0.5761, |
| "step": 93 |
| }, |
| { |
| "epoch": 1.2233956133225021, |
| "grad_norm": 0.8941981850633482, |
| "learning_rate": 7.48232709151145e-05, |
| "loss": 0.5695, |
| "step": 94 |
| }, |
| { |
| "epoch": 1.2363931762794476, |
| "grad_norm": 0.820917113713941, |
| "learning_rate": 7.464101615137756e-05, |
| "loss": 0.5805, |
| "step": 95 |
| }, |
| { |
| "epoch": 1.2493907392363932, |
| "grad_norm": 0.9174411331061607, |
| "learning_rate": 7.445583834532546e-05, |
| "loss": 0.5649, |
| "step": 96 |
| }, |
| { |
| "epoch": 1.2623883021933389, |
| "grad_norm": 0.6500765370181957, |
| "learning_rate": 7.426775312243986e-05, |
| "loss": 0.5758, |
| "step": 97 |
| }, |
| { |
| "epoch": 1.2753858651502843, |
| "grad_norm": 0.763425847678293, |
| "learning_rate": 7.407677635353308e-05, |
| "loss": 0.574, |
| "step": 98 |
| }, |
| { |
| "epoch": 1.28838342810723, |
| "grad_norm": 0.8352567885058773, |
| "learning_rate": 7.388292415340888e-05, |
| "loss": 0.5685, |
| "step": 99 |
| }, |
| { |
| "epoch": 1.3013809910641756, |
| "grad_norm": 0.4826298499702842, |
| "learning_rate": 7.368621287950264e-05, |
| "loss": 0.5645, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.314378554021121, |
| "grad_norm": 0.5903830880665961, |
| "learning_rate": 7.348665913050115e-05, |
| "loss": 0.5602, |
| "step": 101 |
| }, |
| { |
| "epoch": 1.3273761169780667, |
| "grad_norm": 0.790277966015901, |
| "learning_rate": 7.328427974494201e-05, |
| "loss": 0.5676, |
| "step": 102 |
| }, |
| { |
| "epoch": 1.340373679935012, |
| "grad_norm": 0.7512617541518178, |
| "learning_rate": 7.307909179979274e-05, |
| "loss": 0.5648, |
| "step": 103 |
| }, |
| { |
| "epoch": 1.3533712428919578, |
| "grad_norm": 0.5851295705571637, |
| "learning_rate": 7.28711126090098e-05, |
| "loss": 0.5651, |
| "step": 104 |
| }, |
| { |
| "epoch": 1.3663688058489034, |
| "grad_norm": 0.41913986382054075, |
| "learning_rate": 7.266035972207773e-05, |
| "loss": 0.5572, |
| "step": 105 |
| }, |
| { |
| "epoch": 1.3793663688058488, |
| "grad_norm": 0.3907177567251715, |
| "learning_rate": 7.24468509225281e-05, |
| "loss": 0.5689, |
| "step": 106 |
| }, |
| { |
| "epoch": 1.3923639317627945, |
| "grad_norm": 0.510204926248653, |
| "learning_rate": 7.223060422643914e-05, |
| "loss": 0.5685, |
| "step": 107 |
| }, |
| { |
| "epoch": 1.40536149471974, |
| "grad_norm": 0.4592923316364611, |
| "learning_rate": 7.201163788091536e-05, |
| "loss": 0.5669, |
| "step": 108 |
| }, |
| { |
| "epoch": 1.4183590576766856, |
| "grad_norm": 0.6513129221318326, |
| "learning_rate": 7.178997036254799e-05, |
| "loss": 0.5681, |
| "step": 109 |
| }, |
| { |
| "epoch": 1.4313566206336312, |
| "grad_norm": 0.7772542110813966, |
| "learning_rate": 7.156562037585576e-05, |
| "loss": 0.5661, |
| "step": 110 |
| }, |
| { |
| "epoch": 1.4443541835905767, |
| "grad_norm": 0.6513876710701023, |
| "learning_rate": 7.133860685170665e-05, |
| "loss": 0.5736, |
| "step": 111 |
| }, |
| { |
| "epoch": 1.4573517465475223, |
| "grad_norm": 0.576315717747176, |
| "learning_rate": 7.110894894572056e-05, |
| "loss": 0.5641, |
| "step": 112 |
| }, |
| { |
| "epoch": 1.470349309504468, |
| "grad_norm": 0.5234871451279655, |
| "learning_rate": 7.087666603665284e-05, |
| "loss": 0.5606, |
| "step": 113 |
| }, |
| { |
| "epoch": 1.4833468724614134, |
| "grad_norm": 0.5433209657397478, |
| "learning_rate": 7.064177772475912e-05, |
| "loss": 0.5677, |
| "step": 114 |
| }, |
| { |
| "epoch": 1.496344435418359, |
| "grad_norm": 0.5010015862310137, |
| "learning_rate": 7.040430383014146e-05, |
| "loss": 0.5653, |
| "step": 115 |
| }, |
| { |
| "epoch": 1.5093419983753047, |
| "grad_norm": 0.48832775679905743, |
| "learning_rate": 7.016426439107586e-05, |
| "loss": 0.5644, |
| "step": 116 |
| }, |
| { |
| "epoch": 1.5223395613322501, |
| "grad_norm": 0.6320240467713527, |
| "learning_rate": 6.992167966232143e-05, |
| "loss": 0.5567, |
| "step": 117 |
| }, |
| { |
| "epoch": 1.5353371242891958, |
| "grad_norm": 0.6482838878203654, |
| "learning_rate": 6.967657011341126e-05, |
| "loss": 0.5749, |
| "step": 118 |
| }, |
| { |
| "epoch": 1.5483346872461414, |
| "grad_norm": 0.4549530011344001, |
| "learning_rate": 6.942895642692527e-05, |
| "loss": 0.561, |
| "step": 119 |
| }, |
| { |
| "epoch": 1.5613322502030869, |
| "grad_norm": 0.32795847922028915, |
| "learning_rate": 6.917885949674483e-05, |
| "loss": 0.565, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.5743298131600325, |
| "grad_norm": 0.4131817811198266, |
| "learning_rate": 6.892630042628988e-05, |
| "loss": 0.5559, |
| "step": 121 |
| }, |
| { |
| "epoch": 1.5873273761169782, |
| "grad_norm": 0.34923043241455753, |
| "learning_rate": 6.867130052673806e-05, |
| "loss": 0.5627, |
| "step": 122 |
| }, |
| { |
| "epoch": 1.6003249390739236, |
| "grad_norm": 0.4012122766550806, |
| "learning_rate": 6.841388131522656e-05, |
| "loss": 0.561, |
| "step": 123 |
| }, |
| { |
| "epoch": 1.6133225020308692, |
| "grad_norm": 0.45749696954329366, |
| "learning_rate": 6.815406451303647e-05, |
| "loss": 0.5619, |
| "step": 124 |
| }, |
| { |
| "epoch": 1.626320064987815, |
| "grad_norm": 0.3572180927283862, |
| "learning_rate": 6.789187204375981e-05, |
| "loss": 0.5637, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.6393176279447603, |
| "grad_norm": 0.37728781845068143, |
| "learning_rate": 6.762732603144978e-05, |
| "loss": 0.5605, |
| "step": 126 |
| }, |
| { |
| "epoch": 1.652315190901706, |
| "grad_norm": 0.4270428246291779, |
| "learning_rate": 6.736044879875373e-05, |
| "loss": 0.5536, |
| "step": 127 |
| }, |
| { |
| "epoch": 1.6653127538586516, |
| "grad_norm": 0.299705291536536, |
| "learning_rate": 6.709126286502965e-05, |
| "loss": 0.5623, |
| "step": 128 |
| }, |
| { |
| "epoch": 1.678310316815597, |
| "grad_norm": 0.36624782812777956, |
| "learning_rate": 6.681979094444596e-05, |
| "loss": 0.5569, |
| "step": 129 |
| }, |
| { |
| "epoch": 1.6913078797725425, |
| "grad_norm": 0.33387655040953185, |
| "learning_rate": 6.654605594406486e-05, |
| "loss": 0.5575, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.7043054427294884, |
| "grad_norm": 0.38850086520685223, |
| "learning_rate": 6.627008096190938e-05, |
| "loss": 0.5573, |
| "step": 131 |
| }, |
| { |
| "epoch": 1.7173030056864338, |
| "grad_norm": 0.412039261196286, |
| "learning_rate": 6.59918892850144e-05, |
| "loss": 0.5598, |
| "step": 132 |
| }, |
| { |
| "epoch": 1.7303005686433792, |
| "grad_norm": 0.4639064578164096, |
| "learning_rate": 6.571150438746157e-05, |
| "loss": 0.5573, |
| "step": 133 |
| }, |
| { |
| "epoch": 1.743298131600325, |
| "grad_norm": 0.569316518837733, |
| "learning_rate": 6.542894992839873e-05, |
| "loss": 0.5525, |
| "step": 134 |
| }, |
| { |
| "epoch": 1.7562956945572705, |
| "grad_norm": 0.7768980117231334, |
| "learning_rate": 6.514424975004329e-05, |
| "loss": 0.5569, |
| "step": 135 |
| }, |
| { |
| "epoch": 1.769293257514216, |
| "grad_norm": 0.8690174213866712, |
| "learning_rate": 6.48574278756706e-05, |
| "loss": 0.5581, |
| "step": 136 |
| }, |
| { |
| "epoch": 1.7822908204711616, |
| "grad_norm": 0.7189834805856271, |
| "learning_rate": 6.456850850758673e-05, |
| "loss": 0.5523, |
| "step": 137 |
| }, |
| { |
| "epoch": 1.7952883834281073, |
| "grad_norm": 0.42461801914103414, |
| "learning_rate": 6.427751602508628e-05, |
| "loss": 0.5513, |
| "step": 138 |
| }, |
| { |
| "epoch": 1.8082859463850527, |
| "grad_norm": 0.3638089421301393, |
| "learning_rate": 6.398447498239527e-05, |
| "loss": 0.5597, |
| "step": 139 |
| }, |
| { |
| "epoch": 1.8212835093419983, |
| "grad_norm": 0.4699533771325944, |
| "learning_rate": 6.368941010659921e-05, |
| "loss": 0.5582, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.834281072298944, |
| "grad_norm": 0.4419813323602383, |
| "learning_rate": 6.339234629555655e-05, |
| "loss": 0.5561, |
| "step": 141 |
| }, |
| { |
| "epoch": 1.8472786352558894, |
| "grad_norm": 0.36271241045794583, |
| "learning_rate": 6.309330861579786e-05, |
| "loss": 0.5584, |
| "step": 142 |
| }, |
| { |
| "epoch": 1.860276198212835, |
| "grad_norm": 0.40965930158743397, |
| "learning_rate": 6.279232230041065e-05, |
| "loss": 0.5536, |
| "step": 143 |
| }, |
| { |
| "epoch": 1.8732737611697807, |
| "grad_norm": 0.40198532643300333, |
| "learning_rate": 6.248941274691017e-05, |
| "loss": 0.5511, |
| "step": 144 |
| }, |
| { |
| "epoch": 1.8862713241267262, |
| "grad_norm": 0.46299430742363373, |
| "learning_rate": 6.218460551509636e-05, |
| "loss": 0.5471, |
| "step": 145 |
| }, |
| { |
| "epoch": 1.8992688870836718, |
| "grad_norm": 0.5215127675687683, |
| "learning_rate": 6.18779263248971e-05, |
| "loss": 0.5553, |
| "step": 146 |
| }, |
| { |
| "epoch": 1.9122664500406175, |
| "grad_norm": 0.45729570755510884, |
| "learning_rate": 6.156940105419785e-05, |
| "loss": 0.5491, |
| "step": 147 |
| }, |
| { |
| "epoch": 1.925264012997563, |
| "grad_norm": 0.3015515408877481, |
| "learning_rate": 6.125905573665824e-05, |
| "loss": 0.5512, |
| "step": 148 |
| }, |
| { |
| "epoch": 1.9382615759545085, |
| "grad_norm": 0.4496510050196559, |
| "learning_rate": 6.094691655951512e-05, |
| "loss": 0.5537, |
| "step": 149 |
| }, |
| { |
| "epoch": 1.9512591389114542, |
| "grad_norm": 0.5782505006225216, |
| "learning_rate": 6.063300986137297e-05, |
| "loss": 0.5519, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.9642567018683996, |
| "grad_norm": 0.48811843212065303, |
| "learning_rate": 6.0317362129981375e-05, |
| "loss": 0.5519, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.9772542648253453, |
| "grad_norm": 0.3524549635448169, |
| "learning_rate": 6.000000000000001e-05, |
| "loss": 0.5555, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.990251827782291, |
| "grad_norm": 0.3499895937576908, |
| "learning_rate": 5.968095025075114e-05, |
| "loss": 0.5571, |
| "step": 153 |
| }, |
| { |
| "epoch": 2.0048740861088548, |
| "grad_norm": 0.354088344538403, |
| "learning_rate": 5.936023980395997e-05, |
| "loss": 0.5401, |
| "step": 154 |
| }, |
| { |
| "epoch": 2.0178716490658, |
| "grad_norm": 0.4330890417017954, |
| "learning_rate": 5.903789572148295e-05, |
| "loss": 0.5114, |
| "step": 155 |
| }, |
| { |
| "epoch": 2.0308692120227456, |
| "grad_norm": 0.45299756805326713, |
| "learning_rate": 5.871394520302432e-05, |
| "loss": 0.5027, |
| "step": 156 |
| }, |
| { |
| "epoch": 2.0438667749796915, |
| "grad_norm": 0.36185553048035746, |
| "learning_rate": 5.838841558384091e-05, |
| "loss": 0.5004, |
| "step": 157 |
| }, |
| { |
| "epoch": 2.056864337936637, |
| "grad_norm": 0.4094243535630425, |
| "learning_rate": 5.806133433243558e-05, |
| "loss": 0.5089, |
| "step": 158 |
| }, |
| { |
| "epoch": 2.0698619008935824, |
| "grad_norm": 0.4119895854921856, |
| "learning_rate": 5.7732729048239444e-05, |
| "loss": 0.5041, |
| "step": 159 |
| }, |
| { |
| "epoch": 2.0828594638505282, |
| "grad_norm": 0.49951764317825964, |
| "learning_rate": 5.740262745928293e-05, |
| "loss": 0.5083, |
| "step": 160 |
| }, |
| { |
| "epoch": 2.0958570268074737, |
| "grad_norm": 0.6492588353532962, |
| "learning_rate": 5.707105741985615e-05, |
| "loss": 0.5089, |
| "step": 161 |
| }, |
| { |
| "epoch": 2.108854589764419, |
| "grad_norm": 0.8172546815280193, |
| "learning_rate": 5.673804690815845e-05, |
| "loss": 0.5125, |
| "step": 162 |
| }, |
| { |
| "epoch": 2.121852152721365, |
| "grad_norm": 0.8291933191610376, |
| "learning_rate": 5.6403624023937614e-05, |
| "loss": 0.5042, |
| "step": 163 |
| }, |
| { |
| "epoch": 2.1348497156783104, |
| "grad_norm": 0.6400143821799831, |
| "learning_rate": 5.606781698611879e-05, |
| "loss": 0.5019, |
| "step": 164 |
| }, |
| { |
| "epoch": 2.147847278635256, |
| "grad_norm": 0.5436979640183732, |
| "learning_rate": 5.573065413042333e-05, |
| "loss": 0.5045, |
| "step": 165 |
| }, |
| { |
| "epoch": 2.1608448415922012, |
| "grad_norm": 0.4492438740170881, |
| "learning_rate": 5.5392163906977835e-05, |
| "loss": 0.5067, |
| "step": 166 |
| }, |
| { |
| "epoch": 2.173842404549147, |
| "grad_norm": 0.40317590644042245, |
| "learning_rate": 5.505237487791343e-05, |
| "loss": 0.5041, |
| "step": 167 |
| }, |
| { |
| "epoch": 2.1868399675060926, |
| "grad_norm": 0.4137259663946232, |
| "learning_rate": 5.471131571495574e-05, |
| "loss": 0.5008, |
| "step": 168 |
| }, |
| { |
| "epoch": 2.199837530463038, |
| "grad_norm": 0.39271569786675564, |
| "learning_rate": 5.4369015197005506e-05, |
| "loss": 0.5109, |
| "step": 169 |
| }, |
| { |
| "epoch": 2.212835093419984, |
| "grad_norm": 0.3609357892178772, |
| "learning_rate": 5.4025502207710184e-05, |
| "loss": 0.5056, |
| "step": 170 |
| }, |
| { |
| "epoch": 2.2258326563769293, |
| "grad_norm": 0.32273394848970144, |
| "learning_rate": 5.368080573302676e-05, |
| "loss": 0.5004, |
| "step": 171 |
| }, |
| { |
| "epoch": 2.2388302193338747, |
| "grad_norm": 0.3902918387294514, |
| "learning_rate": 5.333495485877583e-05, |
| "loss": 0.5015, |
| "step": 172 |
| }, |
| { |
| "epoch": 2.2518277822908206, |
| "grad_norm": 0.3615272576113914, |
| "learning_rate": 5.298797876818735e-05, |
| "loss": 0.501, |
| "step": 173 |
| }, |
| { |
| "epoch": 2.264825345247766, |
| "grad_norm": 0.3092837009560793, |
| "learning_rate": 5.263990673943811e-05, |
| "loss": 0.5013, |
| "step": 174 |
| }, |
| { |
| "epoch": 2.2778229082047114, |
| "grad_norm": 0.3221337815028307, |
| "learning_rate": 5.229076814318122e-05, |
| "loss": 0.5023, |
| "step": 175 |
| }, |
| { |
| "epoch": 2.2908204711616573, |
| "grad_norm": 0.31876856314125945, |
| "learning_rate": 5.194059244006779e-05, |
| "loss": 0.4954, |
| "step": 176 |
| }, |
| { |
| "epoch": 2.3038180341186028, |
| "grad_norm": 0.3471396479864812, |
| "learning_rate": 5.158940917826099e-05, |
| "loss": 0.5041, |
| "step": 177 |
| }, |
| { |
| "epoch": 2.316815597075548, |
| "grad_norm": 0.30244403952200244, |
| "learning_rate": 5.123724799094279e-05, |
| "loss": 0.4968, |
| "step": 178 |
| }, |
| { |
| "epoch": 2.329813160032494, |
| "grad_norm": 0.27081780367373653, |
| "learning_rate": 5.088413859381341e-05, |
| "loss": 0.495, |
| "step": 179 |
| }, |
| { |
| "epoch": 2.3428107229894395, |
| "grad_norm": 0.26698926546704715, |
| "learning_rate": 5.053011078258397e-05, |
| "loss": 0.4986, |
| "step": 180 |
| }, |
| { |
| "epoch": 2.355808285946385, |
| "grad_norm": 0.2434254540157736, |
| "learning_rate": 5.017519443046226e-05, |
| "loss": 0.5059, |
| "step": 181 |
| }, |
| { |
| "epoch": 2.368805848903331, |
| "grad_norm": 0.2649836381754834, |
| "learning_rate": 4.981941948563197e-05, |
| "loss": 0.4939, |
| "step": 182 |
| }, |
| { |
| "epoch": 2.381803411860276, |
| "grad_norm": 0.24587524499861643, |
| "learning_rate": 4.94628159687257e-05, |
| "loss": 0.4955, |
| "step": 183 |
| }, |
| { |
| "epoch": 2.3948009748172217, |
| "grad_norm": 0.2099952601611583, |
| "learning_rate": 4.9105413970291747e-05, |
| "loss": 0.4998, |
| "step": 184 |
| }, |
| { |
| "epoch": 2.4077985377741675, |
| "grad_norm": 0.30733517181039643, |
| "learning_rate": 4.874724364825504e-05, |
| "loss": 0.5051, |
| "step": 185 |
| }, |
| { |
| "epoch": 2.420796100731113, |
| "grad_norm": 0.23936641127353223, |
| "learning_rate": 4.8388335225372416e-05, |
| "loss": 0.5042, |
| "step": 186 |
| }, |
| { |
| "epoch": 2.4337936636880584, |
| "grad_norm": 0.24807141319926213, |
| "learning_rate": 4.802871898668237e-05, |
| "loss": 0.5015, |
| "step": 187 |
| }, |
| { |
| "epoch": 2.4467912266450043, |
| "grad_norm": 0.23764312316041677, |
| "learning_rate": 4.7668425276949546e-05, |
| "loss": 0.5047, |
| "step": 188 |
| }, |
| { |
| "epoch": 2.4597887896019497, |
| "grad_norm": 0.19839288524791437, |
| "learning_rate": 4.730748449810429e-05, |
| "loss": 0.4945, |
| "step": 189 |
| }, |
| { |
| "epoch": 2.472786352558895, |
| "grad_norm": 0.21343457304874502, |
| "learning_rate": 4.694592710667723e-05, |
| "loss": 0.4948, |
| "step": 190 |
| }, |
| { |
| "epoch": 2.4857839155158405, |
| "grad_norm": 0.23293583325724027, |
| "learning_rate": 4.658378361122936e-05, |
| "loss": 0.4986, |
| "step": 191 |
| }, |
| { |
| "epoch": 2.4987814784727864, |
| "grad_norm": 0.21561162546636445, |
| "learning_rate": 4.622108456977773e-05, |
| "loss": 0.5027, |
| "step": 192 |
| }, |
| { |
| "epoch": 2.511779041429732, |
| "grad_norm": 0.20220078631632762, |
| "learning_rate": 4.585786058721687e-05, |
| "loss": 0.4992, |
| "step": 193 |
| }, |
| { |
| "epoch": 2.5247766043866777, |
| "grad_norm": 0.19015117259585296, |
| "learning_rate": 4.549414231273633e-05, |
| "loss": 0.5019, |
| "step": 194 |
| }, |
| { |
| "epoch": 2.537774167343623, |
| "grad_norm": 0.17888404685442738, |
| "learning_rate": 4.512996043723453e-05, |
| "loss": 0.5009, |
| "step": 195 |
| }, |
| { |
| "epoch": 2.5507717303005686, |
| "grad_norm": 0.1891627608715842, |
| "learning_rate": 4.476534569072895e-05, |
| "loss": 0.5005, |
| "step": 196 |
| }, |
| { |
| "epoch": 2.563769293257514, |
| "grad_norm": 0.1960672578073265, |
| "learning_rate": 4.440032883976318e-05, |
| "loss": 0.4979, |
| "step": 197 |
| }, |
| { |
| "epoch": 2.57676685621446, |
| "grad_norm": 0.1907240431557997, |
| "learning_rate": 4.403494068481074e-05, |
| "loss": 0.4965, |
| "step": 198 |
| }, |
| { |
| "epoch": 2.5897644191714053, |
| "grad_norm": 0.21281767640684754, |
| "learning_rate": 4.3669212057676145e-05, |
| "loss": 0.5035, |
| "step": 199 |
| }, |
| { |
| "epoch": 2.602761982128351, |
| "grad_norm": 0.20034978185083457, |
| "learning_rate": 4.33031738188933e-05, |
| "loss": 0.493, |
| "step": 200 |
| }, |
| { |
| "epoch": 2.6157595450852966, |
| "grad_norm": 0.18482405886366973, |
| "learning_rate": 4.293685685512142e-05, |
| "loss": 0.4979, |
| "step": 201 |
| }, |
| { |
| "epoch": 2.628757108042242, |
| "grad_norm": 0.20689788018303523, |
| "learning_rate": 4.257029207653881e-05, |
| "loss": 0.4968, |
| "step": 202 |
| }, |
| { |
| "epoch": 2.6417546709991875, |
| "grad_norm": 0.20468709904511792, |
| "learning_rate": 4.220351041423462e-05, |
| "loss": 0.496, |
| "step": 203 |
| }, |
| { |
| "epoch": 2.6547522339561334, |
| "grad_norm": 0.19708900779722158, |
| "learning_rate": 4.183654281759888e-05, |
| "loss": 0.5059, |
| "step": 204 |
| }, |
| { |
| "epoch": 2.667749796913079, |
| "grad_norm": 0.1694897534384111, |
| "learning_rate": 4.1469420251710905e-05, |
| "loss": 0.5001, |
| "step": 205 |
| }, |
| { |
| "epoch": 2.680747359870024, |
| "grad_norm": 0.2060507236034366, |
| "learning_rate": 4.110217369472649e-05, |
| "loss": 0.5042, |
| "step": 206 |
| }, |
| { |
| "epoch": 2.69374492282697, |
| "grad_norm": 0.21981117761095914, |
| "learning_rate": 4.07348341352639e-05, |
| "loss": 0.4925, |
| "step": 207 |
| }, |
| { |
| "epoch": 2.7067424857839155, |
| "grad_norm": 0.15869661704467267, |
| "learning_rate": 4.0367432569789065e-05, |
| "loss": 0.4925, |
| "step": 208 |
| }, |
| { |
| "epoch": 2.719740048740861, |
| "grad_norm": 0.17554453102907516, |
| "learning_rate": 4e-05, |
| "loss": 0.4898, |
| "step": 209 |
| }, |
| { |
| "epoch": 2.732737611697807, |
| "grad_norm": 0.20109162996025604, |
| "learning_rate": 3.963256743021095e-05, |
| "loss": 0.4935, |
| "step": 210 |
| }, |
| { |
| "epoch": 2.7457351746547523, |
| "grad_norm": 0.16965164550202852, |
| "learning_rate": 3.92651658647361e-05, |
| "loss": 0.4926, |
| "step": 211 |
| }, |
| { |
| "epoch": 2.7587327376116977, |
| "grad_norm": 0.19107787557510944, |
| "learning_rate": 3.889782630527353e-05, |
| "loss": 0.5003, |
| "step": 212 |
| }, |
| { |
| "epoch": 2.7717303005686436, |
| "grad_norm": 0.23600064349601493, |
| "learning_rate": 3.853057974828911e-05, |
| "loss": 0.5025, |
| "step": 213 |
| }, |
| { |
| "epoch": 2.784727863525589, |
| "grad_norm": 0.1938736742926907, |
| "learning_rate": 3.816345718240113e-05, |
| "loss": 0.496, |
| "step": 214 |
| }, |
| { |
| "epoch": 2.7977254264825344, |
| "grad_norm": 0.18674334250731434, |
| "learning_rate": 3.779648958576538e-05, |
| "loss": 0.4964, |
| "step": 215 |
| }, |
| { |
| "epoch": 2.81072298943948, |
| "grad_norm": 0.17897442021815202, |
| "learning_rate": 3.74297079234612e-05, |
| "loss": 0.4993, |
| "step": 216 |
| }, |
| { |
| "epoch": 2.8237205523964257, |
| "grad_norm": 0.1454966971656852, |
| "learning_rate": 3.706314314487859e-05, |
| "loss": 0.5024, |
| "step": 217 |
| }, |
| { |
| "epoch": 2.836718115353371, |
| "grad_norm": 0.17254506937932373, |
| "learning_rate": 3.669682618110671e-05, |
| "loss": 0.496, |
| "step": 218 |
| }, |
| { |
| "epoch": 2.849715678310317, |
| "grad_norm": 0.17898500506822662, |
| "learning_rate": 3.6330787942323855e-05, |
| "loss": 0.4971, |
| "step": 219 |
| }, |
| { |
| "epoch": 2.8627132412672625, |
| "grad_norm": 0.1775497414815859, |
| "learning_rate": 3.5965059315189274e-05, |
| "loss": 0.4994, |
| "step": 220 |
| }, |
| { |
| "epoch": 2.875710804224208, |
| "grad_norm": 0.15173916947706884, |
| "learning_rate": 3.559967116023683e-05, |
| "loss": 0.501, |
| "step": 221 |
| }, |
| { |
| "epoch": 2.8887083671811533, |
| "grad_norm": 0.2019676520697475, |
| "learning_rate": 3.523465430927106e-05, |
| "loss": 0.4975, |
| "step": 222 |
| }, |
| { |
| "epoch": 2.901705930138099, |
| "grad_norm": 0.17820583073365817, |
| "learning_rate": 3.4870039562765475e-05, |
| "loss": 0.497, |
| "step": 223 |
| }, |
| { |
| "epoch": 2.9147034930950446, |
| "grad_norm": 0.17230973288348897, |
| "learning_rate": 3.4505857687263675e-05, |
| "loss": 0.4918, |
| "step": 224 |
| }, |
| { |
| "epoch": 2.9277010560519905, |
| "grad_norm": 0.17781379922544957, |
| "learning_rate": 3.414213941278314e-05, |
| "loss": 0.5009, |
| "step": 225 |
| }, |
| { |
| "epoch": 2.940698619008936, |
| "grad_norm": 0.17434301641204888, |
| "learning_rate": 3.377891543022229e-05, |
| "loss": 0.4967, |
| "step": 226 |
| }, |
| { |
| "epoch": 2.9536961819658814, |
| "grad_norm": 0.15878938609241405, |
| "learning_rate": 3.341621638877064e-05, |
| "loss": 0.5022, |
| "step": 227 |
| }, |
| { |
| "epoch": 2.966693744922827, |
| "grad_norm": 0.1877350498093743, |
| "learning_rate": 3.305407289332279e-05, |
| "loss": 0.4943, |
| "step": 228 |
| }, |
| { |
| "epoch": 2.9796913078797727, |
| "grad_norm": 0.16153562862339216, |
| "learning_rate": 3.269251550189573e-05, |
| "loss": 0.4952, |
| "step": 229 |
| }, |
| { |
| "epoch": 2.992688870836718, |
| "grad_norm": 0.1636698530309882, |
| "learning_rate": 3.2331574723050474e-05, |
| "loss": 0.5036, |
| "step": 230 |
| }, |
| { |
| "epoch": 3.007311129163282, |
| "grad_norm": 0.24083959413053807, |
| "learning_rate": 3.197128101331764e-05, |
| "loss": 0.4787, |
| "step": 231 |
| }, |
| { |
| "epoch": 3.0203086921202273, |
| "grad_norm": 0.20882743227454145, |
| "learning_rate": 3.161166477462759e-05, |
| "loss": 0.4527, |
| "step": 232 |
| }, |
| { |
| "epoch": 3.033306255077173, |
| "grad_norm": 0.2562598902235666, |
| "learning_rate": 3.125275635174497e-05, |
| "loss": 0.4566, |
| "step": 233 |
| }, |
| { |
| "epoch": 3.0463038180341186, |
| "grad_norm": 0.29932911926131295, |
| "learning_rate": 3.089458602970828e-05, |
| "loss": 0.4492, |
| "step": 234 |
| }, |
| { |
| "epoch": 3.059301380991064, |
| "grad_norm": 0.22296216539718972, |
| "learning_rate": 3.0537184031274306e-05, |
| "loss": 0.452, |
| "step": 235 |
| }, |
| { |
| "epoch": 3.07229894394801, |
| "grad_norm": 0.265839442067331, |
| "learning_rate": 3.0180580514368037e-05, |
| "loss": 0.4591, |
| "step": 236 |
| }, |
| { |
| "epoch": 3.0852965069049554, |
| "grad_norm": 0.26317399079157505, |
| "learning_rate": 2.9824805569537747e-05, |
| "loss": 0.4485, |
| "step": 237 |
| }, |
| { |
| "epoch": 3.098294069861901, |
| "grad_norm": 0.21663346321522212, |
| "learning_rate": 2.9469889217416045e-05, |
| "loss": 0.4556, |
| "step": 238 |
| }, |
| { |
| "epoch": 3.1112916328188467, |
| "grad_norm": 0.2650899931767454, |
| "learning_rate": 2.9115861406186593e-05, |
| "loss": 0.4537, |
| "step": 239 |
| }, |
| { |
| "epoch": 3.124289195775792, |
| "grad_norm": 0.2506262975062407, |
| "learning_rate": 2.8762752009057232e-05, |
| "loss": 0.4535, |
| "step": 240 |
| }, |
| { |
| "epoch": 3.1372867587327375, |
| "grad_norm": 0.22119457110883312, |
| "learning_rate": 2.841059082173902e-05, |
| "loss": 0.4502, |
| "step": 241 |
| }, |
| { |
| "epoch": 3.150284321689683, |
| "grad_norm": 0.23862545493558096, |
| "learning_rate": 2.805940755993223e-05, |
| "loss": 0.4537, |
| "step": 242 |
| }, |
| { |
| "epoch": 3.163281884646629, |
| "grad_norm": 0.18800051564766848, |
| "learning_rate": 2.770923185681878e-05, |
| "loss": 0.4534, |
| "step": 243 |
| }, |
| { |
| "epoch": 3.1762794476035743, |
| "grad_norm": 0.21073908239980146, |
| "learning_rate": 2.7360093260561904e-05, |
| "loss": 0.4537, |
| "step": 244 |
| }, |
| { |
| "epoch": 3.1892770105605197, |
| "grad_norm": 0.19564030664878582, |
| "learning_rate": 2.7012021231812666e-05, |
| "loss": 0.4533, |
| "step": 245 |
| }, |
| { |
| "epoch": 3.2022745735174656, |
| "grad_norm": 0.1689855239948926, |
| "learning_rate": 2.6665045141224193e-05, |
| "loss": 0.451, |
| "step": 246 |
| }, |
| { |
| "epoch": 3.215272136474411, |
| "grad_norm": 0.20951707110954032, |
| "learning_rate": 2.6319194266973256e-05, |
| "loss": 0.4557, |
| "step": 247 |
| }, |
| { |
| "epoch": 3.2282696994313564, |
| "grad_norm": 0.16999772257768134, |
| "learning_rate": 2.597449779228983e-05, |
| "loss": 0.4468, |
| "step": 248 |
| }, |
| { |
| "epoch": 3.2412672623883023, |
| "grad_norm": 0.19094834892704782, |
| "learning_rate": 2.563098480299451e-05, |
| "loss": 0.4554, |
| "step": 249 |
| }, |
| { |
| "epoch": 3.2542648253452477, |
| "grad_norm": 0.16246873328527864, |
| "learning_rate": 2.5288684285044283e-05, |
| "loss": 0.4549, |
| "step": 250 |
| }, |
| { |
| "epoch": 3.267262388302193, |
| "grad_norm": 0.1825793103224904, |
| "learning_rate": 2.4947625122086585e-05, |
| "loss": 0.4543, |
| "step": 251 |
| }, |
| { |
| "epoch": 3.280259951259139, |
| "grad_norm": 0.15892372670398716, |
| "learning_rate": 2.460783609302218e-05, |
| "loss": 0.4575, |
| "step": 252 |
| }, |
| { |
| "epoch": 3.2932575142160845, |
| "grad_norm": 0.17673710311378757, |
| "learning_rate": 2.4269345869576676e-05, |
| "loss": 0.452, |
| "step": 253 |
| }, |
| { |
| "epoch": 3.30625507717303, |
| "grad_norm": 0.15259667013543174, |
| "learning_rate": 2.393218301388123e-05, |
| "loss": 0.4507, |
| "step": 254 |
| }, |
| { |
| "epoch": 3.319252640129976, |
| "grad_norm": 0.15618013871442282, |
| "learning_rate": 2.35963759760624e-05, |
| "loss": 0.4483, |
| "step": 255 |
| }, |
| { |
| "epoch": 3.332250203086921, |
| "grad_norm": 0.1383646033070645, |
| "learning_rate": 2.3261953091841553e-05, |
| "loss": 0.4466, |
| "step": 256 |
| }, |
| { |
| "epoch": 3.3452477660438666, |
| "grad_norm": 0.15025307639697422, |
| "learning_rate": 2.2928942580143855e-05, |
| "loss": 0.457, |
| "step": 257 |
| }, |
| { |
| "epoch": 3.3582453290008125, |
| "grad_norm": 0.12942060560157578, |
| "learning_rate": 2.2597372540717083e-05, |
| "loss": 0.4494, |
| "step": 258 |
| }, |
| { |
| "epoch": 3.371242891957758, |
| "grad_norm": 0.13761592476330478, |
| "learning_rate": 2.226727095176057e-05, |
| "loss": 0.4577, |
| "step": 259 |
| }, |
| { |
| "epoch": 3.3842404549147034, |
| "grad_norm": 0.12368831090048717, |
| "learning_rate": 2.1938665667564435e-05, |
| "loss": 0.45, |
| "step": 260 |
| }, |
| { |
| "epoch": 3.3972380178716493, |
| "grad_norm": 0.13980977768107444, |
| "learning_rate": 2.1611584416159106e-05, |
| "loss": 0.4601, |
| "step": 261 |
| }, |
| { |
| "epoch": 3.4102355808285947, |
| "grad_norm": 0.1292981782906244, |
| "learning_rate": 2.1286054796975696e-05, |
| "loss": 0.4544, |
| "step": 262 |
| }, |
| { |
| "epoch": 3.42323314378554, |
| "grad_norm": 0.13081878013901876, |
| "learning_rate": 2.096210427851706e-05, |
| "loss": 0.4499, |
| "step": 263 |
| }, |
| { |
| "epoch": 3.436230706742486, |
| "grad_norm": 0.13234583952511578, |
| "learning_rate": 2.063976019604006e-05, |
| "loss": 0.4492, |
| "step": 264 |
| }, |
| { |
| "epoch": 3.4492282696994314, |
| "grad_norm": 0.13707720449565938, |
| "learning_rate": 2.0319049749248876e-05, |
| "loss": 0.4576, |
| "step": 265 |
| }, |
| { |
| "epoch": 3.462225832656377, |
| "grad_norm": 0.1275609686612589, |
| "learning_rate": 2.0000000000000012e-05, |
| "loss": 0.4478, |
| "step": 266 |
| }, |
| { |
| "epoch": 3.4752233956133223, |
| "grad_norm": 0.13858936276545247, |
| "learning_rate": 1.9682637870018638e-05, |
| "loss": 0.4503, |
| "step": 267 |
| }, |
| { |
| "epoch": 3.488220958570268, |
| "grad_norm": 0.13526985297489078, |
| "learning_rate": 1.9366990138627054e-05, |
| "loss": 0.451, |
| "step": 268 |
| }, |
| { |
| "epoch": 3.5012185215272136, |
| "grad_norm": 0.13143298791548755, |
| "learning_rate": 1.9053083440484887e-05, |
| "loss": 0.4469, |
| "step": 269 |
| }, |
| { |
| "epoch": 3.5142160844841595, |
| "grad_norm": 0.13651218753852146, |
| "learning_rate": 1.8740944263341773e-05, |
| "loss": 0.4538, |
| "step": 270 |
| }, |
| { |
| "epoch": 3.527213647441105, |
| "grad_norm": 0.13559937030356303, |
| "learning_rate": 1.8430598945802156e-05, |
| "loss": 0.4578, |
| "step": 271 |
| }, |
| { |
| "epoch": 3.5402112103980503, |
| "grad_norm": 0.13345629350619237, |
| "learning_rate": 1.8122073675102935e-05, |
| "loss": 0.4501, |
| "step": 272 |
| }, |
| { |
| "epoch": 3.5532087733549957, |
| "grad_norm": 0.13156084596113124, |
| "learning_rate": 1.781539448490365e-05, |
| "loss": 0.4508, |
| "step": 273 |
| }, |
| { |
| "epoch": 3.5662063363119416, |
| "grad_norm": 0.12230522417142756, |
| "learning_rate": 1.7510587253089842e-05, |
| "loss": 0.4564, |
| "step": 274 |
| }, |
| { |
| "epoch": 3.579203899268887, |
| "grad_norm": 0.12733451660742465, |
| "learning_rate": 1.7207677699589355e-05, |
| "loss": 0.4562, |
| "step": 275 |
| }, |
| { |
| "epoch": 3.592201462225833, |
| "grad_norm": 0.13918819865169338, |
| "learning_rate": 1.690669138420215e-05, |
| "loss": 0.4539, |
| "step": 276 |
| }, |
| { |
| "epoch": 3.6051990251827783, |
| "grad_norm": 0.12910395860504983, |
| "learning_rate": 1.6607653704443457e-05, |
| "loss": 0.4549, |
| "step": 277 |
| }, |
| { |
| "epoch": 3.618196588139724, |
| "grad_norm": 0.11652161500078773, |
| "learning_rate": 1.6310589893400804e-05, |
| "loss": 0.4495, |
| "step": 278 |
| }, |
| { |
| "epoch": 3.631194151096669, |
| "grad_norm": 0.1435144989657529, |
| "learning_rate": 1.601552501760473e-05, |
| "loss": 0.4615, |
| "step": 279 |
| }, |
| { |
| "epoch": 3.644191714053615, |
| "grad_norm": 0.11768536430982104, |
| "learning_rate": 1.5722483974913737e-05, |
| "loss": 0.452, |
| "step": 280 |
| }, |
| { |
| "epoch": 3.6571892770105605, |
| "grad_norm": 0.12124441944302719, |
| "learning_rate": 1.5431491492413288e-05, |
| "loss": 0.4509, |
| "step": 281 |
| }, |
| { |
| "epoch": 3.670186839967506, |
| "grad_norm": 0.11219311820617218, |
| "learning_rate": 1.5142572124329418e-05, |
| "loss": 0.4536, |
| "step": 282 |
| }, |
| { |
| "epoch": 3.683184402924452, |
| "grad_norm": 0.11088572665376577, |
| "learning_rate": 1.4855750249956718e-05, |
| "loss": 0.4533, |
| "step": 283 |
| }, |
| { |
| "epoch": 3.6961819658813972, |
| "grad_norm": 0.1114181991679099, |
| "learning_rate": 1.457105007160129e-05, |
| "loss": 0.4583, |
| "step": 284 |
| }, |
| { |
| "epoch": 3.7091795288383427, |
| "grad_norm": 0.1150092625579184, |
| "learning_rate": 1.4288495612538427e-05, |
| "loss": 0.4546, |
| "step": 285 |
| }, |
| { |
| "epoch": 3.722177091795288, |
| "grad_norm": 0.11256641012147124, |
| "learning_rate": 1.4008110714985623e-05, |
| "loss": 0.4562, |
| "step": 286 |
| }, |
| { |
| "epoch": 3.735174654752234, |
| "grad_norm": 0.1121146716270174, |
| "learning_rate": 1.3729919038090627e-05, |
| "loss": 0.4531, |
| "step": 287 |
| }, |
| { |
| "epoch": 3.7481722177091794, |
| "grad_norm": 0.12240062452155943, |
| "learning_rate": 1.3453944055935151e-05, |
| "loss": 0.4554, |
| "step": 288 |
| }, |
| { |
| "epoch": 3.7611697806661253, |
| "grad_norm": 0.10662890450993669, |
| "learning_rate": 1.3180209055554043e-05, |
| "loss": 0.4557, |
| "step": 289 |
| }, |
| { |
| "epoch": 3.7741673436230707, |
| "grad_norm": 0.11472028819426605, |
| "learning_rate": 1.2908737134970367e-05, |
| "loss": 0.4568, |
| "step": 290 |
| }, |
| { |
| "epoch": 3.787164906580016, |
| "grad_norm": 0.11298609292901812, |
| "learning_rate": 1.2639551201246278e-05, |
| "loss": 0.4514, |
| "step": 291 |
| }, |
| { |
| "epoch": 3.8001624695369616, |
| "grad_norm": 0.10976020682705631, |
| "learning_rate": 1.2372673968550229e-05, |
| "loss": 0.4476, |
| "step": 292 |
| }, |
| { |
| "epoch": 3.8131600324939074, |
| "grad_norm": 0.10531475963763727, |
| "learning_rate": 1.2108127956240186e-05, |
| "loss": 0.4471, |
| "step": 293 |
| }, |
| { |
| "epoch": 3.826157595450853, |
| "grad_norm": 0.10623689348961035, |
| "learning_rate": 1.1845935486963546e-05, |
| "loss": 0.4515, |
| "step": 294 |
| }, |
| { |
| "epoch": 3.8391551584077988, |
| "grad_norm": 0.10834285056205656, |
| "learning_rate": 1.158611868477344e-05, |
| "loss": 0.4484, |
| "step": 295 |
| }, |
| { |
| "epoch": 3.852152721364744, |
| "grad_norm": 0.10958894542102338, |
| "learning_rate": 1.1328699473261957e-05, |
| "loss": 0.446, |
| "step": 296 |
| }, |
| { |
| "epoch": 3.8651502843216896, |
| "grad_norm": 0.10719668132712344, |
| "learning_rate": 1.107369957371013e-05, |
| "loss": 0.4587, |
| "step": 297 |
| }, |
| { |
| "epoch": 3.878147847278635, |
| "grad_norm": 0.10865202402459116, |
| "learning_rate": 1.0821140503255174e-05, |
| "loss": 0.4478, |
| "step": 298 |
| }, |
| { |
| "epoch": 3.891145410235581, |
| "grad_norm": 0.1075674174283554, |
| "learning_rate": 1.0571043573074737e-05, |
| "loss": 0.4474, |
| "step": 299 |
| }, |
| { |
| "epoch": 3.9041429731925263, |
| "grad_norm": 0.10420773620004607, |
| "learning_rate": 1.0323429886588743e-05, |
| "loss": 0.4558, |
| "step": 300 |
| }, |
| { |
| "epoch": 3.917140536149472, |
| "grad_norm": 0.10602592166217871, |
| "learning_rate": 1.0078320337678584e-05, |
| "loss": 0.4472, |
| "step": 301 |
| }, |
| { |
| "epoch": 3.9301380991064176, |
| "grad_norm": 0.11214585082716343, |
| "learning_rate": 9.835735608924155e-06, |
| "loss": 0.4532, |
| "step": 302 |
| }, |
| { |
| "epoch": 3.943135662063363, |
| "grad_norm": 0.10680829021211037, |
| "learning_rate": 9.595696169858542e-06, |
| "loss": 0.452, |
| "step": 303 |
| }, |
| { |
| "epoch": 3.9561332250203085, |
| "grad_norm": 0.10193849061969358, |
| "learning_rate": 9.358222275240884e-06, |
| "loss": 0.4548, |
| "step": 304 |
| }, |
| { |
| "epoch": 3.9691307879772544, |
| "grad_norm": 0.10621117615673921, |
| "learning_rate": 9.123333963347166e-06, |
| "loss": 0.4568, |
| "step": 305 |
| }, |
| { |
| "epoch": 3.9821283509342, |
| "grad_norm": 0.10611037774239646, |
| "learning_rate": 8.89105105427945e-06, |
| "loss": 0.4503, |
| "step": 306 |
| }, |
| { |
| "epoch": 3.9951259138911457, |
| "grad_norm": 0.10224681565036052, |
| "learning_rate": 8.661393148293355e-06, |
| "loss": 0.4514, |
| "step": 307 |
| }, |
| { |
| "epoch": 4.0097481722177095, |
| "grad_norm": 0.1629992310965872, |
| "learning_rate": 8.434379624144261e-06, |
| "loss": 0.4384, |
| "step": 308 |
| }, |
| { |
| "epoch": 4.022745735174655, |
| "grad_norm": 0.16305297766108903, |
| "learning_rate": 8.210029637452016e-06, |
| "loss": 0.4219, |
| "step": 309 |
| }, |
| { |
| "epoch": 4.0357432981316, |
| "grad_norm": 0.1152257478387255, |
| "learning_rate": 7.988362119084642e-06, |
| "loss": 0.423, |
| "step": 310 |
| }, |
| { |
| "epoch": 4.048740861088546, |
| "grad_norm": 0.12619317639889938, |
| "learning_rate": 7.769395773560874e-06, |
| "loss": 0.4265, |
| "step": 311 |
| }, |
| { |
| "epoch": 4.061738424045491, |
| "grad_norm": 0.1605893684483608, |
| "learning_rate": 7.553149077471915e-06, |
| "loss": 0.4293, |
| "step": 312 |
| }, |
| { |
| "epoch": 4.074735987002437, |
| "grad_norm": 0.16787469942852876, |
| "learning_rate": 7.3396402779222845e-06, |
| "loss": 0.4269, |
| "step": 313 |
| }, |
| { |
| "epoch": 4.087733549959383, |
| "grad_norm": 0.13639739623886823, |
| "learning_rate": 7.128887390990198e-06, |
| "loss": 0.4203, |
| "step": 314 |
| }, |
| { |
| "epoch": 4.100731112916328, |
| "grad_norm": 0.12896797685517175, |
| "learning_rate": 6.9209082002072725e-06, |
| "loss": 0.4237, |
| "step": 315 |
| }, |
| { |
| "epoch": 4.113728675873274, |
| "grad_norm": 0.13126789348210258, |
| "learning_rate": 6.715720255058e-06, |
| "loss": 0.4323, |
| "step": 316 |
| }, |
| { |
| "epoch": 4.126726238830219, |
| "grad_norm": 0.13882914474015806, |
| "learning_rate": 6.513340869498859e-06, |
| "loss": 0.4232, |
| "step": 317 |
| }, |
| { |
| "epoch": 4.139723801787165, |
| "grad_norm": 0.13742788716133153, |
| "learning_rate": 6.313787120497376e-06, |
| "loss": 0.427, |
| "step": 318 |
| }, |
| { |
| "epoch": 4.15272136474411, |
| "grad_norm": 0.12658794759546255, |
| "learning_rate": 6.117075846591123e-06, |
| "loss": 0.422, |
| "step": 319 |
| }, |
| { |
| "epoch": 4.1657189277010565, |
| "grad_norm": 0.11656720107903619, |
| "learning_rate": 5.923223646466923e-06, |
| "loss": 0.424, |
| "step": 320 |
| }, |
| { |
| "epoch": 4.178716490658002, |
| "grad_norm": 0.12017679947679329, |
| "learning_rate": 5.732246877560146e-06, |
| "loss": 0.4239, |
| "step": 321 |
| }, |
| { |
| "epoch": 4.191714053614947, |
| "grad_norm": 0.1248953898447142, |
| "learning_rate": 5.5441616546745646e-06, |
| "loss": 0.4247, |
| "step": 322 |
| }, |
| { |
| "epoch": 4.204711616571893, |
| "grad_norm": 0.11492392711495635, |
| "learning_rate": 5.358983848622452e-06, |
| "loss": 0.423, |
| "step": 323 |
| }, |
| { |
| "epoch": 4.217709179528838, |
| "grad_norm": 0.10979437544809782, |
| "learning_rate": 5.176729084885508e-06, |
| "loss": 0.426, |
| "step": 324 |
| }, |
| { |
| "epoch": 4.230706742485784, |
| "grad_norm": 0.10996774248170614, |
| "learning_rate": 4.99741274229625e-06, |
| "loss": 0.424, |
| "step": 325 |
| }, |
| { |
| "epoch": 4.24370430544273, |
| "grad_norm": 0.11864930806086794, |
| "learning_rate": 4.821049951740442e-06, |
| "loss": 0.4236, |
| "step": 326 |
| }, |
| { |
| "epoch": 4.256701868399675, |
| "grad_norm": 0.11614244217733552, |
| "learning_rate": 4.647655594880225e-06, |
| "loss": 0.4264, |
| "step": 327 |
| }, |
| { |
| "epoch": 4.269699431356621, |
| "grad_norm": 0.10489433778288049, |
| "learning_rate": 4.4772443028985004e-06, |
| "loss": 0.4306, |
| "step": 328 |
| }, |
| { |
| "epoch": 4.282696994313566, |
| "grad_norm": 0.1045337845950868, |
| "learning_rate": 4.3098304552642385e-06, |
| "loss": 0.4279, |
| "step": 329 |
| }, |
| { |
| "epoch": 4.295694557270512, |
| "grad_norm": 0.10425269207053385, |
| "learning_rate": 4.1454281785191995e-06, |
| "loss": 0.4219, |
| "step": 330 |
| }, |
| { |
| "epoch": 4.308692120227457, |
| "grad_norm": 0.10799333991055605, |
| "learning_rate": 3.984051345085855e-06, |
| "loss": 0.4239, |
| "step": 331 |
| }, |
| { |
| "epoch": 4.3216896831844025, |
| "grad_norm": 0.11546875761746016, |
| "learning_rate": 3.825713572096903e-06, |
| "loss": 0.4188, |
| "step": 332 |
| }, |
| { |
| "epoch": 4.334687246141349, |
| "grad_norm": 0.10158288013636599, |
| "learning_rate": 3.6704282202461515e-06, |
| "loss": 0.4243, |
| "step": 333 |
| }, |
| { |
| "epoch": 4.347684809098294, |
| "grad_norm": 0.09998970634420433, |
| "learning_rate": 3.518208392661184e-06, |
| "loss": 0.4221, |
| "step": 334 |
| }, |
| { |
| "epoch": 4.36068237205524, |
| "grad_norm": 0.1072651386764742, |
| "learning_rate": 3.3690669337977e-06, |
| "loss": 0.4204, |
| "step": 335 |
| }, |
| { |
| "epoch": 4.373679935012185, |
| "grad_norm": 0.10189852488748347, |
| "learning_rate": 3.2230164283556918e-06, |
| "loss": 0.4218, |
| "step": 336 |
| }, |
| { |
| "epoch": 4.3866774979691305, |
| "grad_norm": 0.09855467315288223, |
| "learning_rate": 3.080069200217497e-06, |
| "loss": 0.424, |
| "step": 337 |
| }, |
| { |
| "epoch": 4.399675060926076, |
| "grad_norm": 0.09678471475059303, |
| "learning_rate": 2.9402373114079295e-06, |
| "loss": 0.4237, |
| "step": 338 |
| }, |
| { |
| "epoch": 4.412672623883022, |
| "grad_norm": 0.09641330522520603, |
| "learning_rate": 2.803532561076492e-06, |
| "loss": 0.4294, |
| "step": 339 |
| }, |
| { |
| "epoch": 4.425670186839968, |
| "grad_norm": 0.09993981323574717, |
| "learning_rate": 2.669966484501716e-06, |
| "loss": 0.4291, |
| "step": 340 |
| }, |
| { |
| "epoch": 4.438667749796913, |
| "grad_norm": 0.09491135655205166, |
| "learning_rate": 2.5395503521178143e-06, |
| "loss": 0.425, |
| "step": 341 |
| }, |
| { |
| "epoch": 4.451665312753859, |
| "grad_norm": 0.10027652165084279, |
| "learning_rate": 2.4122951685636674e-06, |
| "loss": 0.4287, |
| "step": 342 |
| }, |
| { |
| "epoch": 4.464662875710804, |
| "grad_norm": 0.09768195784704334, |
| "learning_rate": 2.2882116717542634e-06, |
| "loss": 0.4297, |
| "step": 343 |
| }, |
| { |
| "epoch": 4.477660438667749, |
| "grad_norm": 0.09320352715058745, |
| "learning_rate": 2.1673103319746146e-06, |
| "loss": 0.422, |
| "step": 344 |
| }, |
| { |
| "epoch": 4.490658001624696, |
| "grad_norm": 0.09297954183991879, |
| "learning_rate": 2.049601350996233e-06, |
| "loss": 0.4216, |
| "step": 345 |
| }, |
| { |
| "epoch": 4.503655564581641, |
| "grad_norm": 0.09580363639852763, |
| "learning_rate": 1.93509466121633e-06, |
| "loss": 0.4221, |
| "step": 346 |
| }, |
| { |
| "epoch": 4.516653127538587, |
| "grad_norm": 0.09572555821764618, |
| "learning_rate": 1.8237999248197002e-06, |
| "loss": 0.4261, |
| "step": 347 |
| }, |
| { |
| "epoch": 4.529650690495532, |
| "grad_norm": 0.09729341717065641, |
| "learning_rate": 1.7157265329634354e-06, |
| "loss": 0.422, |
| "step": 348 |
| }, |
| { |
| "epoch": 4.5426482534524775, |
| "grad_norm": 0.09262430982141642, |
| "learning_rate": 1.6108836049844434e-06, |
| "loss": 0.4275, |
| "step": 349 |
| }, |
| { |
| "epoch": 4.555645816409423, |
| "grad_norm": 0.09643151222343373, |
| "learning_rate": 1.5092799876299835e-06, |
| "loss": 0.4318, |
| "step": 350 |
| }, |
| { |
| "epoch": 4.568643379366369, |
| "grad_norm": 0.09552392662018869, |
| "learning_rate": 1.4109242543111834e-06, |
| "loss": 0.4262, |
| "step": 351 |
| }, |
| { |
| "epoch": 4.581640942323315, |
| "grad_norm": 0.0934634073485968, |
| "learning_rate": 1.3158247043795735e-06, |
| "loss": 0.4191, |
| "step": 352 |
| }, |
| { |
| "epoch": 4.59463850528026, |
| "grad_norm": 0.09051362955312056, |
| "learning_rate": 1.2239893624267852e-06, |
| "loss": 0.4249, |
| "step": 353 |
| }, |
| { |
| "epoch": 4.6076360682372055, |
| "grad_norm": 0.10311548227178086, |
| "learning_rate": 1.1354259776074472e-06, |
| "loss": 0.421, |
| "step": 354 |
| }, |
| { |
| "epoch": 4.620633631194151, |
| "grad_norm": 0.09034380750409249, |
| "learning_rate": 1.050142022985292e-06, |
| "loss": 0.4234, |
| "step": 355 |
| }, |
| { |
| "epoch": 4.633631194151096, |
| "grad_norm": 0.0912218320639531, |
| "learning_rate": 9.681446949025752e-07, |
| "loss": 0.4231, |
| "step": 356 |
| }, |
| { |
| "epoch": 4.646628757108042, |
| "grad_norm": 0.0907014935264679, |
| "learning_rate": 8.89440912372832e-07, |
| "loss": 0.4162, |
| "step": 357 |
| }, |
| { |
| "epoch": 4.659626320064988, |
| "grad_norm": 0.08963535898951595, |
| "learning_rate": 8.140373164970428e-07, |
| "loss": 0.4286, |
| "step": 358 |
| }, |
| { |
| "epoch": 4.6726238830219335, |
| "grad_norm": 0.09182311609539749, |
| "learning_rate": 7.419402699032852e-07, |
| "loss": 0.4213, |
| "step": 359 |
| }, |
| { |
| "epoch": 4.685621445978879, |
| "grad_norm": 0.08836898405911321, |
| "learning_rate": 6.731558562097995e-07, |
| "loss": 0.4247, |
| "step": 360 |
| }, |
| { |
| "epoch": 4.698619008935824, |
| "grad_norm": 0.09028610838105593, |
| "learning_rate": 6.076898795116792e-07, |
| "loss": 0.4253, |
| "step": 361 |
| }, |
| { |
| "epoch": 4.71161657189277, |
| "grad_norm": 0.09100823747148952, |
| "learning_rate": 5.455478638911071e-07, |
| "loss": 0.4251, |
| "step": 362 |
| }, |
| { |
| "epoch": 4.724614134849716, |
| "grad_norm": 0.09029025365560213, |
| "learning_rate": 4.867350529512261e-07, |
| "loss": 0.4187, |
| "step": 363 |
| }, |
| { |
| "epoch": 4.737611697806662, |
| "grad_norm": 0.08743205850314692, |
| "learning_rate": 4.3125640937368373e-07, |
| "loss": 0.4254, |
| "step": 364 |
| }, |
| { |
| "epoch": 4.750609260763607, |
| "grad_norm": 0.0922555367394928, |
| "learning_rate": 3.791166144998704e-07, |
| "loss": 0.423, |
| "step": 365 |
| }, |
| { |
| "epoch": 4.763606823720552, |
| "grad_norm": 0.08845599991505329, |
| "learning_rate": 3.3032006793590977e-07, |
| "loss": 0.4224, |
| "step": 366 |
| }, |
| { |
| "epoch": 4.776604386677498, |
| "grad_norm": 0.0915393101514765, |
| "learning_rate": 2.848708871814054e-07, |
| "loss": 0.4255, |
| "step": 367 |
| }, |
| { |
| "epoch": 4.789601949634443, |
| "grad_norm": 0.0888246620842322, |
| "learning_rate": 2.4277290728202063e-07, |
| "loss": 0.4202, |
| "step": 368 |
| }, |
| { |
| "epoch": 4.802599512591389, |
| "grad_norm": 0.0912092493664706, |
| "learning_rate": 2.040296805058528e-07, |
| "loss": 0.4316, |
| "step": 369 |
| }, |
| { |
| "epoch": 4.815597075548335, |
| "grad_norm": 0.09110426289157889, |
| "learning_rate": 1.6864447604370004e-07, |
| "loss": 0.43, |
| "step": 370 |
| }, |
| { |
| "epoch": 4.8285946385052805, |
| "grad_norm": 0.08633872492794921, |
| "learning_rate": 1.3662027973320614e-07, |
| "loss": 0.4311, |
| "step": 371 |
| }, |
| { |
| "epoch": 4.841592201462226, |
| "grad_norm": 0.08888780145070413, |
| "learning_rate": 1.0795979380690657e-07, |
| "loss": 0.4182, |
| "step": 372 |
| }, |
| { |
| "epoch": 4.854589764419171, |
| "grad_norm": 0.08984098123092447, |
| "learning_rate": 8.266543666421544e-08, |
| "loss": 0.4172, |
| "step": 373 |
| }, |
| { |
| "epoch": 4.867587327376117, |
| "grad_norm": 0.08852973851348196, |
| "learning_rate": 6.073934266735303e-08, |
| "loss": 0.4234, |
| "step": 374 |
| }, |
| { |
| "epoch": 4.880584890333062, |
| "grad_norm": 0.08776688037991925, |
| "learning_rate": 4.218336196125439e-08, |
| "loss": 0.4268, |
| "step": 375 |
| }, |
| { |
| "epoch": 4.8935824532900085, |
| "grad_norm": 0.08708562125303575, |
| "learning_rate": 2.699906031745414e-08, |
| "loss": 0.4234, |
| "step": 376 |
| }, |
| { |
| "epoch": 4.906580016246954, |
| "grad_norm": 0.08866044810066966, |
| "learning_rate": 1.5187719001943378e-08, |
| "loss": 0.4217, |
| "step": 377 |
| }, |
| { |
| "epoch": 4.919577579203899, |
| "grad_norm": 0.08852375482148439, |
| "learning_rate": 6.750334667091629e-09, |
| "loss": 0.4235, |
| "step": 378 |
| }, |
| { |
| "epoch": 4.932575142160845, |
| "grad_norm": 0.08832697038072729, |
| "learning_rate": 1.6876192675052695e-09, |
| "loss": 0.4274, |
| "step": 379 |
| }, |
| { |
| "epoch": 4.94557270511779, |
| "grad_norm": 0.08596221358605577, |
| "learning_rate": 0.0, |
| "loss": 0.4228, |
| "step": 380 |
| }, |
| { |
| "epoch": 4.94557270511779, |
| "step": 380, |
| "total_flos": 9.778530867524665e+18, |
| "train_loss": 0.5321214740997866, |
| "train_runtime": 38508.3509, |
| "train_samples_per_second": 5.112, |
| "train_steps_per_second": 0.01 |
| } |
| ], |
| "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": 9.778530867524665e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
