| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.0006535922081351315, | |
| "eval_steps": 500, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.307184416270263e-06, | |
| "grad_norm": 6.268255710601807, | |
| "learning_rate": 0.0, | |
| "loss": 7.0486, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 2.614368832540526e-06, | |
| "grad_norm": 5.648464202880859, | |
| "learning_rate": 5e-06, | |
| "loss": 6.6548, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 3.921553248810789e-06, | |
| "grad_norm": 3.917975902557373, | |
| "learning_rate": 1e-05, | |
| "loss": 5.9005, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 5.228737665081052e-06, | |
| "grad_norm": 5.040616989135742, | |
| "learning_rate": 1.5e-05, | |
| "loss": 6.4287, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 6.535922081351315e-06, | |
| "grad_norm": 4.18554162979126, | |
| "learning_rate": 2e-05, | |
| "loss": 5.8283, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 7.843106497621578e-06, | |
| "grad_norm": 3.837583303451538, | |
| "learning_rate": 2.5e-05, | |
| "loss": 5.8711, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 9.150290913891842e-06, | |
| "grad_norm": 3.9937901496887207, | |
| "learning_rate": 3e-05, | |
| "loss": 6.2397, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.0457475330162104e-05, | |
| "grad_norm": 6.323297023773193, | |
| "learning_rate": 3.5e-05, | |
| "loss": 6.0184, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764659746432368e-05, | |
| "grad_norm": 4.541579246520996, | |
| "learning_rate": 4e-05, | |
| "loss": 4.9289, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.307184416270263e-05, | |
| "grad_norm": 4.3712358474731445, | |
| "learning_rate": 4.5e-05, | |
| "loss": 5.5431, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4379028578972894e-05, | |
| "grad_norm": 5.535464763641357, | |
| "learning_rate": 5e-05, | |
| "loss": 5.4152, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.5686212995243156e-05, | |
| "grad_norm": 4.411280632019043, | |
| "learning_rate": 4.9994994994994995e-05, | |
| "loss": 5.0227, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6993397411513418e-05, | |
| "grad_norm": 8.777021408081055, | |
| "learning_rate": 4.9989989989989995e-05, | |
| "loss": 4.628, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.8300581827783683e-05, | |
| "grad_norm": 11.268546104431152, | |
| "learning_rate": 4.998498498498499e-05, | |
| "loss": 4.4529, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.9607766244053945e-05, | |
| "grad_norm": 11.081679344177246, | |
| "learning_rate": 4.997997997997999e-05, | |
| "loss": 3.7927, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0914950660324208e-05, | |
| "grad_norm": 17.39815330505371, | |
| "learning_rate": 4.9974974974974974e-05, | |
| "loss": 3.8354, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.222213507659447e-05, | |
| "grad_norm": 19.696117401123047, | |
| "learning_rate": 4.9969969969969974e-05, | |
| "loss": 2.6296, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.3529319492864735e-05, | |
| "grad_norm": 9.491320610046387, | |
| "learning_rate": 4.996496496496497e-05, | |
| "loss": 2.34, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.4836503909134997e-05, | |
| "grad_norm": 8.10347843170166, | |
| "learning_rate": 4.995995995995996e-05, | |
| "loss": 2.8097, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.614368832540526e-05, | |
| "grad_norm": 10.142881393432617, | |
| "learning_rate": 4.995495495495496e-05, | |
| "loss": 2.418, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.745087274167552e-05, | |
| "grad_norm": 5.666007041931152, | |
| "learning_rate": 4.994994994994995e-05, | |
| "loss": 2.4814, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.8758057157945787e-05, | |
| "grad_norm": 2.6253504753112793, | |
| "learning_rate": 4.994494494494495e-05, | |
| "loss": 3.0029, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 3.006524157421605e-05, | |
| "grad_norm": 1.9413844347000122, | |
| "learning_rate": 4.9939939939939945e-05, | |
| "loss": 2.7044, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 3.137242599048631e-05, | |
| "grad_norm": 2.3702781200408936, | |
| "learning_rate": 4.993493493493494e-05, | |
| "loss": 3.2683, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.267961040675658e-05, | |
| "grad_norm": 1.8796427249908447, | |
| "learning_rate": 4.992992992992993e-05, | |
| "loss": 2.6591, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.3986794823026836e-05, | |
| "grad_norm": 1.8859280347824097, | |
| "learning_rate": 4.9924924924924924e-05, | |
| "loss": 2.6567, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.52939792392971e-05, | |
| "grad_norm": 1.8021355867385864, | |
| "learning_rate": 4.9919919919919924e-05, | |
| "loss": 2.0643, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.660116365556737e-05, | |
| "grad_norm": 3.3398263454437256, | |
| "learning_rate": 4.991491491491492e-05, | |
| "loss": 1.9806, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.7908348071837625e-05, | |
| "grad_norm": 1.7556695938110352, | |
| "learning_rate": 4.9909909909909917e-05, | |
| "loss": 2.5847, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.921553248810789e-05, | |
| "grad_norm": 2.361534833908081, | |
| "learning_rate": 4.990490490490491e-05, | |
| "loss": 3.1982, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.052271690437815e-05, | |
| "grad_norm": 2.1116466522216797, | |
| "learning_rate": 4.98998998998999e-05, | |
| "loss": 2.1447, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 4.1829901320648415e-05, | |
| "grad_norm": 2.76759672164917, | |
| "learning_rate": 4.9894894894894896e-05, | |
| "loss": 2.2138, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.313708573691868e-05, | |
| "grad_norm": 1.744004726409912, | |
| "learning_rate": 4.988988988988989e-05, | |
| "loss": 2.6433, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 4.444427015318894e-05, | |
| "grad_norm": 1.946826457977295, | |
| "learning_rate": 4.988488488488489e-05, | |
| "loss": 2.814, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.5751454569459205e-05, | |
| "grad_norm": 1.7588024139404297, | |
| "learning_rate": 4.987987987987988e-05, | |
| "loss": 1.9231, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.705863898572947e-05, | |
| "grad_norm": 1.920486330986023, | |
| "learning_rate": 4.9874874874874874e-05, | |
| "loss": 2.3058, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.836582340199973e-05, | |
| "grad_norm": 2.172233819961548, | |
| "learning_rate": 4.9869869869869874e-05, | |
| "loss": 1.9368, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 4.9673007818269995e-05, | |
| "grad_norm": 1.9940109252929688, | |
| "learning_rate": 4.986486486486487e-05, | |
| "loss": 2.4671, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 5.0980192234540254e-05, | |
| "grad_norm": 2.1106722354888916, | |
| "learning_rate": 4.985985985985987e-05, | |
| "loss": 1.2266, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 5.228737665081052e-05, | |
| "grad_norm": 1.9987304210662842, | |
| "learning_rate": 4.985485485485486e-05, | |
| "loss": 2.5097, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.3594561067080785e-05, | |
| "grad_norm": 4.258189678192139, | |
| "learning_rate": 4.984984984984985e-05, | |
| "loss": 2.7884, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 5.490174548335104e-05, | |
| "grad_norm": 3.9918763637542725, | |
| "learning_rate": 4.9844844844844846e-05, | |
| "loss": 1.8792, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.620892989962131e-05, | |
| "grad_norm": 2.13405442237854, | |
| "learning_rate": 4.983983983983984e-05, | |
| "loss": 2.4141, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 5.7516114315891574e-05, | |
| "grad_norm": 2.369387626647949, | |
| "learning_rate": 4.983483483483484e-05, | |
| "loss": 2.4503, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.882329873216183e-05, | |
| "grad_norm": 2.4659979343414307, | |
| "learning_rate": 4.982982982982983e-05, | |
| "loss": 1.6704, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 6.01304831484321e-05, | |
| "grad_norm": 2.8049533367156982, | |
| "learning_rate": 4.982482482482483e-05, | |
| "loss": 2.2134, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 6.143766756470236e-05, | |
| "grad_norm": 3.652784824371338, | |
| "learning_rate": 4.9819819819819824e-05, | |
| "loss": 2.4586, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 6.274485198097262e-05, | |
| "grad_norm": 2.76751971244812, | |
| "learning_rate": 4.981481481481482e-05, | |
| "loss": 2.6559, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.405203639724289e-05, | |
| "grad_norm": 2.337362051010132, | |
| "learning_rate": 4.980980980980981e-05, | |
| "loss": 2.6549, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 6.535922081351315e-05, | |
| "grad_norm": 2.5444376468658447, | |
| "learning_rate": 4.98048048048048e-05, | |
| "loss": 2.041, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.666640522978342e-05, | |
| "grad_norm": 2.3144052028656006, | |
| "learning_rate": 4.97997997997998e-05, | |
| "loss": 2.4423, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.797358964605367e-05, | |
| "grad_norm": 2.647606372833252, | |
| "learning_rate": 4.9794794794794796e-05, | |
| "loss": 2.0042, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.928077406232394e-05, | |
| "grad_norm": 2.2047841548919678, | |
| "learning_rate": 4.9789789789789795e-05, | |
| "loss": 2.7494, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 7.05879584785942e-05, | |
| "grad_norm": 2.5686540603637695, | |
| "learning_rate": 4.978478478478479e-05, | |
| "loss": 2.6443, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 7.189514289486447e-05, | |
| "grad_norm": 2.634040594100952, | |
| "learning_rate": 4.977977977977978e-05, | |
| "loss": 2.3489, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 7.320232731113473e-05, | |
| "grad_norm": 3.5904624462127686, | |
| "learning_rate": 4.977477477477478e-05, | |
| "loss": 2.325, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 7.450951172740499e-05, | |
| "grad_norm": 2.7417197227478027, | |
| "learning_rate": 4.976976976976977e-05, | |
| "loss": 2.109, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 7.581669614367525e-05, | |
| "grad_norm": 3.90405011177063, | |
| "learning_rate": 4.976476476476477e-05, | |
| "loss": 1.9777, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 7.712388055994552e-05, | |
| "grad_norm": 3.0904104709625244, | |
| "learning_rate": 4.975975975975976e-05, | |
| "loss": 2.964, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.843106497621578e-05, | |
| "grad_norm": 6.283293724060059, | |
| "learning_rate": 4.975475475475476e-05, | |
| "loss": 2.7203, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.973824939248605e-05, | |
| "grad_norm": 15.482819557189941, | |
| "learning_rate": 4.974974974974975e-05, | |
| "loss": 2.0104, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 8.10454338087563e-05, | |
| "grad_norm": 5.350860118865967, | |
| "learning_rate": 4.9744744744744746e-05, | |
| "loss": 2.4738, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 8.235261822502656e-05, | |
| "grad_norm": 2.5188300609588623, | |
| "learning_rate": 4.9739739739739745e-05, | |
| "loss": 2.3628, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 8.365980264129683e-05, | |
| "grad_norm": 3.6089985370635986, | |
| "learning_rate": 4.973473473473474e-05, | |
| "loss": 2.0068, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 8.49669870575671e-05, | |
| "grad_norm": 4.064090251922607, | |
| "learning_rate": 4.972972972972974e-05, | |
| "loss": 1.9134, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 8.627417147383736e-05, | |
| "grad_norm": 4.058968544006348, | |
| "learning_rate": 4.9724724724724724e-05, | |
| "loss": 2.654, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.758135589010763e-05, | |
| "grad_norm": 4.4842023849487305, | |
| "learning_rate": 4.971971971971972e-05, | |
| "loss": 2.0954, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 8.888854030637788e-05, | |
| "grad_norm": 4.771529674530029, | |
| "learning_rate": 4.971471471471472e-05, | |
| "loss": 2.3041, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 9.019572472264814e-05, | |
| "grad_norm": 5.2841267585754395, | |
| "learning_rate": 4.970970970970971e-05, | |
| "loss": 1.6284, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 9.150290913891841e-05, | |
| "grad_norm": 4.454013824462891, | |
| "learning_rate": 4.970470470470471e-05, | |
| "loss": 2.4258, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 9.281009355518868e-05, | |
| "grad_norm": 4.686120986938477, | |
| "learning_rate": 4.96996996996997e-05, | |
| "loss": 1.7159, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 9.411727797145894e-05, | |
| "grad_norm": 4.039860248565674, | |
| "learning_rate": 4.9694694694694696e-05, | |
| "loss": 2.4571, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 9.542446238772919e-05, | |
| "grad_norm": 6.6892805099487305, | |
| "learning_rate": 4.9689689689689696e-05, | |
| "loss": 1.5466, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 9.673164680399946e-05, | |
| "grad_norm": 4.6148362159729, | |
| "learning_rate": 4.968468468468468e-05, | |
| "loss": 2.1115, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 9.803883122026972e-05, | |
| "grad_norm": 4.417403697967529, | |
| "learning_rate": 4.967967967967968e-05, | |
| "loss": 2.0975, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 9.934601563653999e-05, | |
| "grad_norm": 6.658927917480469, | |
| "learning_rate": 4.9674674674674674e-05, | |
| "loss": 1.4255, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.00010065320005281026, | |
| "grad_norm": 5.162082195281982, | |
| "learning_rate": 4.9669669669669674e-05, | |
| "loss": 2.1406, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.00010196038446908051, | |
| "grad_norm": 3.9566612243652344, | |
| "learning_rate": 4.966466466466467e-05, | |
| "loss": 2.6316, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.00010326756888535077, | |
| "grad_norm": 5.431234836578369, | |
| "learning_rate": 4.965965965965966e-05, | |
| "loss": 1.9831, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.00010457475330162104, | |
| "grad_norm": 7.695541858673096, | |
| "learning_rate": 4.965465465465466e-05, | |
| "loss": 1.3602, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0001058819377178913, | |
| "grad_norm": 5.668361663818359, | |
| "learning_rate": 4.964964964964965e-05, | |
| "loss": 1.2717, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.00010718912213416157, | |
| "grad_norm": 6.2884202003479, | |
| "learning_rate": 4.9644644644644646e-05, | |
| "loss": 2.3994, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.00010849630655043183, | |
| "grad_norm": 5.497302055358887, | |
| "learning_rate": 4.963963963963964e-05, | |
| "loss": 2.0307, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.00010980349096670209, | |
| "grad_norm": 6.879200458526611, | |
| "learning_rate": 4.963463463463464e-05, | |
| "loss": 1.4382, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.00011111067538297235, | |
| "grad_norm": 5.275783538818359, | |
| "learning_rate": 4.962962962962963e-05, | |
| "loss": 1.8855, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.00011241785979924262, | |
| "grad_norm": 5.1828460693359375, | |
| "learning_rate": 4.9624624624624625e-05, | |
| "loss": 2.1653, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.00011372504421551288, | |
| "grad_norm": 4.819062232971191, | |
| "learning_rate": 4.9619619619619624e-05, | |
| "loss": 2.6426, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.00011503222863178315, | |
| "grad_norm": 4.319160461425781, | |
| "learning_rate": 4.961461461461462e-05, | |
| "loss": 2.6315, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.0001163394130480534, | |
| "grad_norm": 5.244660377502441, | |
| "learning_rate": 4.960960960960962e-05, | |
| "loss": 2.4654, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.00011764659746432367, | |
| "grad_norm": 4.433946132659912, | |
| "learning_rate": 4.96046046046046e-05, | |
| "loss": 2.1795, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.00011895378188059393, | |
| "grad_norm": 5.009888648986816, | |
| "learning_rate": 4.95995995995996e-05, | |
| "loss": 1.3696, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.0001202609662968642, | |
| "grad_norm": 4.033069133758545, | |
| "learning_rate": 4.9594594594594596e-05, | |
| "loss": 2.3803, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.00012156815071313446, | |
| "grad_norm": 4.739905834197998, | |
| "learning_rate": 4.958958958958959e-05, | |
| "loss": 1.7027, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.00012287533512940471, | |
| "grad_norm": 5.726681232452393, | |
| "learning_rate": 4.958458458458459e-05, | |
| "loss": 1.0941, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.00012418251954567498, | |
| "grad_norm": 4.398964881896973, | |
| "learning_rate": 4.957957957957958e-05, | |
| "loss": 1.8081, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.00012548970396194525, | |
| "grad_norm": 3.442424774169922, | |
| "learning_rate": 4.957457457457458e-05, | |
| "loss": 2.0943, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.0001267968883782155, | |
| "grad_norm": 3.42536997795105, | |
| "learning_rate": 4.9569569569569574e-05, | |
| "loss": 2.3138, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.00012810407279448578, | |
| "grad_norm": 3.0850987434387207, | |
| "learning_rate": 4.956456456456457e-05, | |
| "loss": 2.1422, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.00012941125721075604, | |
| "grad_norm": 3.327355146408081, | |
| "learning_rate": 4.955955955955956e-05, | |
| "loss": 1.607, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.0001307184416270263, | |
| "grad_norm": 3.0526223182678223, | |
| "learning_rate": 4.955455455455455e-05, | |
| "loss": 1.5995, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.00013202562604329657, | |
| "grad_norm": 3.1441359519958496, | |
| "learning_rate": 4.954954954954955e-05, | |
| "loss": 2.2129, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.00013333281045956684, | |
| "grad_norm": 2.209366798400879, | |
| "learning_rate": 4.9544544544544546e-05, | |
| "loss": 2.2904, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.00013463999487583708, | |
| "grad_norm": 2.4867329597473145, | |
| "learning_rate": 4.953953953953954e-05, | |
| "loss": 0.7869, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.00013594717929210734, | |
| "grad_norm": 2.2363412380218506, | |
| "learning_rate": 4.953453453453454e-05, | |
| "loss": 1.4709, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0001372543637083776, | |
| "grad_norm": 2.6723251342773438, | |
| "learning_rate": 4.952952952952953e-05, | |
| "loss": 2.2871, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.00013856154812464787, | |
| "grad_norm": 1.9363102912902832, | |
| "learning_rate": 4.952452452452453e-05, | |
| "loss": 1.985, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.00013986873254091814, | |
| "grad_norm": 1.7824273109436035, | |
| "learning_rate": 4.951951951951952e-05, | |
| "loss": 1.189, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0001411759169571884, | |
| "grad_norm": 1.8107385635375977, | |
| "learning_rate": 4.951451451451452e-05, | |
| "loss": 1.8943, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.00014248310137345867, | |
| "grad_norm": 2.283268451690674, | |
| "learning_rate": 4.950950950950951e-05, | |
| "loss": 1.7242, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.00014379028578972894, | |
| "grad_norm": 1.7572216987609863, | |
| "learning_rate": 4.95045045045045e-05, | |
| "loss": 1.8113, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0001450974702059992, | |
| "grad_norm": 2.1798532009124756, | |
| "learning_rate": 4.94994994994995e-05, | |
| "loss": 1.9223, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.00014640465462226947, | |
| "grad_norm": 1.4888501167297363, | |
| "learning_rate": 4.9494494494494496e-05, | |
| "loss": 1.9747, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0001477118390385397, | |
| "grad_norm": 1.8382623195648193, | |
| "learning_rate": 4.9489489489489496e-05, | |
| "loss": 2.199, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.00014901902345480997, | |
| "grad_norm": 2.766442060470581, | |
| "learning_rate": 4.948448448448449e-05, | |
| "loss": 1.433, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.00015032620787108024, | |
| "grad_norm": 1.3955047130584717, | |
| "learning_rate": 4.947947947947948e-05, | |
| "loss": 0.8611, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.0001516333922873505, | |
| "grad_norm": 1.3797332048416138, | |
| "learning_rate": 4.9474474474474475e-05, | |
| "loss": 1.8795, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.00015294057670362077, | |
| "grad_norm": 2.196755886077881, | |
| "learning_rate": 4.946946946946947e-05, | |
| "loss": 1.4086, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.00015424776111989103, | |
| "grad_norm": 1.6287676095962524, | |
| "learning_rate": 4.946446446446447e-05, | |
| "loss": 1.0982, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.0001555549455361613, | |
| "grad_norm": 2.1588571071624756, | |
| "learning_rate": 4.945945945945946e-05, | |
| "loss": 2.2011, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.00015686212995243156, | |
| "grad_norm": 1.66507887840271, | |
| "learning_rate": 4.945445445445446e-05, | |
| "loss": 1.6342, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.00015816931436870183, | |
| "grad_norm": 2.086681365966797, | |
| "learning_rate": 4.944944944944945e-05, | |
| "loss": 2.4012, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.0001594764987849721, | |
| "grad_norm": 2.2638509273529053, | |
| "learning_rate": 4.9444444444444446e-05, | |
| "loss": 0.2397, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.00016078368320124236, | |
| "grad_norm": 1.6955903768539429, | |
| "learning_rate": 4.9439439439439446e-05, | |
| "loss": 2.0733, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.0001620908676175126, | |
| "grad_norm": 1.7107908725738525, | |
| "learning_rate": 4.943443443443443e-05, | |
| "loss": 1.6862, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.00016339805203378286, | |
| "grad_norm": 2.117300033569336, | |
| "learning_rate": 4.942942942942943e-05, | |
| "loss": 1.6494, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.00016470523645005313, | |
| "grad_norm": 1.9111508131027222, | |
| "learning_rate": 4.9424424424424425e-05, | |
| "loss": 0.932, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.0001660124208663234, | |
| "grad_norm": 1.934354305267334, | |
| "learning_rate": 4.9419419419419425e-05, | |
| "loss": 1.6937, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.00016731960528259366, | |
| "grad_norm": 2.0459141731262207, | |
| "learning_rate": 4.941441441441442e-05, | |
| "loss": 2.1459, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.00016862678969886393, | |
| "grad_norm": 1.9724762439727783, | |
| "learning_rate": 4.940940940940941e-05, | |
| "loss": 2.2388, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.0001699339741151342, | |
| "grad_norm": 1.900592565536499, | |
| "learning_rate": 4.940440440440441e-05, | |
| "loss": 2.3512, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.00017124115853140446, | |
| "grad_norm": 1.978319764137268, | |
| "learning_rate": 4.93993993993994e-05, | |
| "loss": 1.7366, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.00017254834294767472, | |
| "grad_norm": 1.5513819456100464, | |
| "learning_rate": 4.9394394394394396e-05, | |
| "loss": 1.0159, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.000173855527363945, | |
| "grad_norm": 2.037893772125244, | |
| "learning_rate": 4.938938938938939e-05, | |
| "loss": 1.4329, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.00017516271178021525, | |
| "grad_norm": 1.7013237476348877, | |
| "learning_rate": 4.938438438438439e-05, | |
| "loss": 0.7663, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0001764698961964855, | |
| "grad_norm": 1.8744912147521973, | |
| "learning_rate": 4.937937937937938e-05, | |
| "loss": 1.9305, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.00017777708061275576, | |
| "grad_norm": 1.606590747833252, | |
| "learning_rate": 4.9374374374374375e-05, | |
| "loss": 2.1672, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.00017908426502902602, | |
| "grad_norm": 2.2351033687591553, | |
| "learning_rate": 4.9369369369369375e-05, | |
| "loss": 1.0281, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.0001803914494452963, | |
| "grad_norm": 2.2086477279663086, | |
| "learning_rate": 4.936436436436437e-05, | |
| "loss": 1.6579, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.00018169863386156655, | |
| "grad_norm": 1.6347638368606567, | |
| "learning_rate": 4.935935935935936e-05, | |
| "loss": 1.7343, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.00018300581827783682, | |
| "grad_norm": 1.6986087560653687, | |
| "learning_rate": 4.9354354354354354e-05, | |
| "loss": 2.2753, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.00018431300269410709, | |
| "grad_norm": 1.8067278861999512, | |
| "learning_rate": 4.9349349349349347e-05, | |
| "loss": 0.8229, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.00018562018711037735, | |
| "grad_norm": 1.9648691415786743, | |
| "learning_rate": 4.9344344344344346e-05, | |
| "loss": 1.8146, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.00018692737152664762, | |
| "grad_norm": 1.3609802722930908, | |
| "learning_rate": 4.933933933933934e-05, | |
| "loss": 1.4422, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.00018823455594291788, | |
| "grad_norm": 2.002868890762329, | |
| "learning_rate": 4.933433433433434e-05, | |
| "loss": 0.0777, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.00018954174035918812, | |
| "grad_norm": 1.5004322528839111, | |
| "learning_rate": 4.932932932932933e-05, | |
| "loss": 1.5963, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.00019084892477545839, | |
| "grad_norm": 1.5936583280563354, | |
| "learning_rate": 4.9324324324324325e-05, | |
| "loss": 0.765, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.00019215610919172865, | |
| "grad_norm": 1.2652479410171509, | |
| "learning_rate": 4.9319319319319325e-05, | |
| "loss": 0.9597, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.00019346329360799892, | |
| "grad_norm": 1.7863197326660156, | |
| "learning_rate": 4.931431431431432e-05, | |
| "loss": 0.9391, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.00019477047802426918, | |
| "grad_norm": 1.9188827276229858, | |
| "learning_rate": 4.930930930930931e-05, | |
| "loss": 2.2355, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.00019607766244053945, | |
| "grad_norm": 1.5928704738616943, | |
| "learning_rate": 4.9304304304304304e-05, | |
| "loss": 1.1829, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0001973848468568097, | |
| "grad_norm": 2.145599603652954, | |
| "learning_rate": 4.92992992992993e-05, | |
| "loss": 1.4609, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.00019869203127307998, | |
| "grad_norm": 1.4171509742736816, | |
| "learning_rate": 4.9294294294294296e-05, | |
| "loss": 1.9975, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.00019999921568935024, | |
| "grad_norm": 1.3210326433181763, | |
| "learning_rate": 4.928928928928929e-05, | |
| "loss": 1.0295, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.0002013064001056205, | |
| "grad_norm": 2.013183832168579, | |
| "learning_rate": 4.928428428428429e-05, | |
| "loss": 1.5366, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.00020261358452189078, | |
| "grad_norm": 1.594954490661621, | |
| "learning_rate": 4.927927927927928e-05, | |
| "loss": 1.0073, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.00020392076893816101, | |
| "grad_norm": 2.206082344055176, | |
| "learning_rate": 4.927427427427428e-05, | |
| "loss": 0.9262, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.00020522795335443128, | |
| "grad_norm": 2.40515398979187, | |
| "learning_rate": 4.926926926926927e-05, | |
| "loss": 1.4697, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.00020653513777070155, | |
| "grad_norm": 1.8254011869430542, | |
| "learning_rate": 4.926426426426427e-05, | |
| "loss": 1.8146, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.0002078423221869718, | |
| "grad_norm": 2.0213472843170166, | |
| "learning_rate": 4.925925925925926e-05, | |
| "loss": 3.1616, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.00020914950660324208, | |
| "grad_norm": 1.8707919120788574, | |
| "learning_rate": 4.9254254254254254e-05, | |
| "loss": 1.6477, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.00021045669101951234, | |
| "grad_norm": 1.9849202632904053, | |
| "learning_rate": 4.9249249249249253e-05, | |
| "loss": 1.5499, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.0002117638754357826, | |
| "grad_norm": 1.8460415601730347, | |
| "learning_rate": 4.9244244244244246e-05, | |
| "loss": 1.4391, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.00021307105985205287, | |
| "grad_norm": 1.8488810062408447, | |
| "learning_rate": 4.9239239239239246e-05, | |
| "loss": 1.8629, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.00021437824426832314, | |
| "grad_norm": 1.8770229816436768, | |
| "learning_rate": 4.923423423423424e-05, | |
| "loss": 2.3753, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.0002156854286845934, | |
| "grad_norm": 1.472601056098938, | |
| "learning_rate": 4.922922922922923e-05, | |
| "loss": 0.7564, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.00021699261310086367, | |
| "grad_norm": 2.860130548477173, | |
| "learning_rate": 4.9224224224224225e-05, | |
| "loss": 0.7373, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.0002182997975171339, | |
| "grad_norm": 1.886745572090149, | |
| "learning_rate": 4.921921921921922e-05, | |
| "loss": 1.0192, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.00021960698193340417, | |
| "grad_norm": 1.2749582529067993, | |
| "learning_rate": 4.921421421421422e-05, | |
| "loss": 0.4704, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.00022091416634967444, | |
| "grad_norm": 1.645389199256897, | |
| "learning_rate": 4.920920920920921e-05, | |
| "loss": 1.8075, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.0002222213507659447, | |
| "grad_norm": 1.7299386262893677, | |
| "learning_rate": 4.920420420420421e-05, | |
| "loss": 2.4038, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.00022352853518221497, | |
| "grad_norm": 1.6160026788711548, | |
| "learning_rate": 4.9199199199199204e-05, | |
| "loss": 1.1931, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.00022483571959848524, | |
| "grad_norm": 1.9258168935775757, | |
| "learning_rate": 4.9194194194194196e-05, | |
| "loss": 2.2323, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.0002261429040147555, | |
| "grad_norm": 2.079798936843872, | |
| "learning_rate": 4.9189189189189196e-05, | |
| "loss": 1.9468, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.00022745008843102577, | |
| "grad_norm": 2.0454630851745605, | |
| "learning_rate": 4.918418418418418e-05, | |
| "loss": 1.7755, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.00022875727284729603, | |
| "grad_norm": 1.9173661470413208, | |
| "learning_rate": 4.917917917917918e-05, | |
| "loss": 1.0856, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.0002300644572635663, | |
| "grad_norm": 1.6535844802856445, | |
| "learning_rate": 4.9174174174174175e-05, | |
| "loss": 1.6214, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.00023137164167983656, | |
| "grad_norm": 2.402019739151001, | |
| "learning_rate": 4.916916916916917e-05, | |
| "loss": 1.7331, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.0002326788260961068, | |
| "grad_norm": 1.7244210243225098, | |
| "learning_rate": 4.916416416416417e-05, | |
| "loss": 1.4885, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.00023398601051237707, | |
| "grad_norm": 2.702523946762085, | |
| "learning_rate": 4.915915915915916e-05, | |
| "loss": 2.0403, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.00023529319492864733, | |
| "grad_norm": 2.1688618659973145, | |
| "learning_rate": 4.915415415415416e-05, | |
| "loss": 0.986, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.0002366003793449176, | |
| "grad_norm": 2.3919677734375, | |
| "learning_rate": 4.9149149149149154e-05, | |
| "loss": 1.3291, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.00023790756376118786, | |
| "grad_norm": 1.8715122938156128, | |
| "learning_rate": 4.9144144144144147e-05, | |
| "loss": 1.7018, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.00023921474817745813, | |
| "grad_norm": 2.152589797973633, | |
| "learning_rate": 4.913913913913914e-05, | |
| "loss": 1.8486, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.0002405219325937284, | |
| "grad_norm": 1.5322504043579102, | |
| "learning_rate": 4.913413413413413e-05, | |
| "loss": 1.6896, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.00024182911700999866, | |
| "grad_norm": 1.6208884716033936, | |
| "learning_rate": 4.912912912912913e-05, | |
| "loss": 1.9508, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.00024313630142626893, | |
| "grad_norm": 1.5740910768508911, | |
| "learning_rate": 4.9124124124124125e-05, | |
| "loss": 1.1913, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.00024444348584253916, | |
| "grad_norm": 2.652970790863037, | |
| "learning_rate": 4.9119119119119125e-05, | |
| "loss": 1.5786, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.00024575067025880943, | |
| "grad_norm": 1.5115643739700317, | |
| "learning_rate": 4.911411411411412e-05, | |
| "loss": 2.3382, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.0002470578546750797, | |
| "grad_norm": 2.009474754333496, | |
| "learning_rate": 4.910910910910911e-05, | |
| "loss": 2.2713, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.00024836503909134996, | |
| "grad_norm": 3.25408673286438, | |
| "learning_rate": 4.9104104104104104e-05, | |
| "loss": 1.767, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.0002496722235076202, | |
| "grad_norm": 1.6666245460510254, | |
| "learning_rate": 4.90990990990991e-05, | |
| "loss": 1.1057, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.0002509794079238905, | |
| "grad_norm": 2.5899922847747803, | |
| "learning_rate": 4.90940940940941e-05, | |
| "loss": 1.7851, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.00025228659234016076, | |
| "grad_norm": 1.9656989574432373, | |
| "learning_rate": 4.908908908908909e-05, | |
| "loss": 2.5767, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.000253593776756431, | |
| "grad_norm": 1.7756541967391968, | |
| "learning_rate": 4.908408408408409e-05, | |
| "loss": 1.7306, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.0002549009611727013, | |
| "grad_norm": 1.9504581689834595, | |
| "learning_rate": 4.907907907907908e-05, | |
| "loss": 1.5729, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.00025620814558897155, | |
| "grad_norm": 0.957115888595581, | |
| "learning_rate": 4.9074074074074075e-05, | |
| "loss": 0.277, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.0002575153300052418, | |
| "grad_norm": 1.4767944812774658, | |
| "learning_rate": 4.9069069069069075e-05, | |
| "loss": 1.5084, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.0002588225144215121, | |
| "grad_norm": 1.2738802433013916, | |
| "learning_rate": 4.906406406406407e-05, | |
| "loss": 0.9465, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.00026012969883778235, | |
| "grad_norm": 1.6053591966629028, | |
| "learning_rate": 4.905905905905906e-05, | |
| "loss": 1.3578, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.0002614368832540526, | |
| "grad_norm": 1.664510726928711, | |
| "learning_rate": 4.9054054054054054e-05, | |
| "loss": 2.4543, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0002627440676703229, | |
| "grad_norm": 1.6944023370742798, | |
| "learning_rate": 4.9049049049049054e-05, | |
| "loss": 1.3043, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.00026405125208659315, | |
| "grad_norm": 1.387467861175537, | |
| "learning_rate": 4.904404404404405e-05, | |
| "loss": 1.4816, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.0002653584365028634, | |
| "grad_norm": 1.925346851348877, | |
| "learning_rate": 4.903903903903904e-05, | |
| "loss": 1.6951, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.0002666656209191337, | |
| "grad_norm": 1.4878056049346924, | |
| "learning_rate": 4.903403403403404e-05, | |
| "loss": 1.7805, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.0002679728053354039, | |
| "grad_norm": 2.018648862838745, | |
| "learning_rate": 4.902902902902903e-05, | |
| "loss": 1.4652, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.00026927998975167415, | |
| "grad_norm": 1.8136177062988281, | |
| "learning_rate": 4.902402402402403e-05, | |
| "loss": 1.5613, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.0002705871741679444, | |
| "grad_norm": 2.323335886001587, | |
| "learning_rate": 4.901901901901902e-05, | |
| "loss": 1.6988, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.0002718943585842147, | |
| "grad_norm": 1.7906171083450317, | |
| "learning_rate": 4.901401401401402e-05, | |
| "loss": 1.5934, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.00027320154300048495, | |
| "grad_norm": 1.5515409708023071, | |
| "learning_rate": 4.900900900900901e-05, | |
| "loss": 1.5874, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.0002745087274167552, | |
| "grad_norm": 1.9337890148162842, | |
| "learning_rate": 4.9004004004004004e-05, | |
| "loss": 1.7147, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.0002758159118330255, | |
| "grad_norm": 2.102532148361206, | |
| "learning_rate": 4.8998998998999004e-05, | |
| "loss": 1.7885, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.00027712309624929575, | |
| "grad_norm": 1.6877130270004272, | |
| "learning_rate": 4.8993993993994e-05, | |
| "loss": 1.5048, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.000278430280665566, | |
| "grad_norm": 1.8921681642532349, | |
| "learning_rate": 4.898898898898899e-05, | |
| "loss": 1.8371, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.0002797374650818363, | |
| "grad_norm": 1.810315728187561, | |
| "learning_rate": 4.898398398398399e-05, | |
| "loss": 1.7009, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.00028104464949810654, | |
| "grad_norm": 2.759514331817627, | |
| "learning_rate": 4.8978978978978976e-05, | |
| "loss": 3.1437, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.0002823518339143768, | |
| "grad_norm": 2.4991655349731445, | |
| "learning_rate": 4.8973973973973975e-05, | |
| "loss": 2.3588, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.0002836590183306471, | |
| "grad_norm": 2.547462224960327, | |
| "learning_rate": 4.896896896896897e-05, | |
| "loss": 1.5821, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.00028496620274691734, | |
| "grad_norm": 1.3106446266174316, | |
| "learning_rate": 4.896396396396397e-05, | |
| "loss": 1.4013, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.0002862733871631876, | |
| "grad_norm": 1.6845613718032837, | |
| "learning_rate": 4.895895895895896e-05, | |
| "loss": 1.5136, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.00028758057157945787, | |
| "grad_norm": 1.659991979598999, | |
| "learning_rate": 4.8953953953953954e-05, | |
| "loss": 2.2554, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.00028888775599572814, | |
| "grad_norm": 1.3590683937072754, | |
| "learning_rate": 4.8948948948948954e-05, | |
| "loss": 0.7868, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.0002901949404119984, | |
| "grad_norm": 1.8435338735580444, | |
| "learning_rate": 4.894394394394395e-05, | |
| "loss": 1.8352, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.00029150212482826867, | |
| "grad_norm": 1.6166387796401978, | |
| "learning_rate": 4.893893893893894e-05, | |
| "loss": 1.3868, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.00029280930924453893, | |
| "grad_norm": 2.057384967803955, | |
| "learning_rate": 4.893393393393393e-05, | |
| "loss": 1.898, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.0002941164936608092, | |
| "grad_norm": 1.695014238357544, | |
| "learning_rate": 4.892892892892893e-05, | |
| "loss": 2.2495, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.0002954236780770794, | |
| "grad_norm": 2.4186930656433105, | |
| "learning_rate": 4.8923923923923926e-05, | |
| "loss": 2.1074, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.0002967308624933497, | |
| "grad_norm": 1.5932652950286865, | |
| "learning_rate": 4.891891891891892e-05, | |
| "loss": 1.1245, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.00029803804690961994, | |
| "grad_norm": 1.7992708683013916, | |
| "learning_rate": 4.891391391391392e-05, | |
| "loss": 1.6981, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.0002993452313258902, | |
| "grad_norm": 1.6720223426818848, | |
| "learning_rate": 4.890890890890891e-05, | |
| "loss": 1.499, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.00030065241574216047, | |
| "grad_norm": 1.7174367904663086, | |
| "learning_rate": 4.890390390390391e-05, | |
| "loss": 1.4546, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.00030195960015843074, | |
| "grad_norm": 2.5477240085601807, | |
| "learning_rate": 4.8898898898898904e-05, | |
| "loss": 1.0959, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.000303266784574701, | |
| "grad_norm": 1.551347017288208, | |
| "learning_rate": 4.88938938938939e-05, | |
| "loss": 1.2192, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.00030457396899097127, | |
| "grad_norm": 4.913365364074707, | |
| "learning_rate": 4.888888888888889e-05, | |
| "loss": 1.4216, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.00030588115340724153, | |
| "grad_norm": 2.8688876628875732, | |
| "learning_rate": 4.888388388388388e-05, | |
| "loss": 1.2954, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.0003071883378235118, | |
| "grad_norm": 1.8472936153411865, | |
| "learning_rate": 4.887887887887888e-05, | |
| "loss": 1.6743, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.00030849552223978207, | |
| "grad_norm": 2.6009321212768555, | |
| "learning_rate": 4.8873873873873876e-05, | |
| "loss": 0.8994, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.00030980270665605233, | |
| "grad_norm": 3.2686116695404053, | |
| "learning_rate": 4.8868868868868875e-05, | |
| "loss": 1.89, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.0003111098910723226, | |
| "grad_norm": 2.842278003692627, | |
| "learning_rate": 4.886386386386387e-05, | |
| "loss": 1.3352, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.00031241707548859286, | |
| "grad_norm": 1.6975959539413452, | |
| "learning_rate": 4.885885885885886e-05, | |
| "loss": 2.5637, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.00031372425990486313, | |
| "grad_norm": 1.6205472946166992, | |
| "learning_rate": 4.8853853853853854e-05, | |
| "loss": 1.7702, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.0003150314443211334, | |
| "grad_norm": 1.9137065410614014, | |
| "learning_rate": 4.884884884884885e-05, | |
| "loss": 2.128, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.00031633862873740366, | |
| "grad_norm": 1.6640243530273438, | |
| "learning_rate": 4.884384384384385e-05, | |
| "loss": 2.798, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.0003176458131536739, | |
| "grad_norm": 1.5545439720153809, | |
| "learning_rate": 4.883883883883884e-05, | |
| "loss": 1.9383, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.0003189529975699442, | |
| "grad_norm": 2.110470771789551, | |
| "learning_rate": 4.883383383383384e-05, | |
| "loss": 2.7099, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.00032026018198621446, | |
| "grad_norm": 1.3277205228805542, | |
| "learning_rate": 4.882882882882883e-05, | |
| "loss": 1.6492, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.0003215673664024847, | |
| "grad_norm": 1.317232608795166, | |
| "learning_rate": 4.8823823823823826e-05, | |
| "loss": 1.3825, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.00032287455081875493, | |
| "grad_norm": 1.5808682441711426, | |
| "learning_rate": 4.8818818818818825e-05, | |
| "loss": 1.7761, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.0003241817352350252, | |
| "grad_norm": 1.576471209526062, | |
| "learning_rate": 4.881381381381381e-05, | |
| "loss": 1.6476, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.00032548891965129546, | |
| "grad_norm": 1.5742518901824951, | |
| "learning_rate": 4.880880880880881e-05, | |
| "loss": 1.3871, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.00032679610406756573, | |
| "grad_norm": 1.22970449924469, | |
| "learning_rate": 4.8803803803803804e-05, | |
| "loss": 0.9665, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.000328103288483836, | |
| "grad_norm": 1.9720549583435059, | |
| "learning_rate": 4.87987987987988e-05, | |
| "loss": 1.5671, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.00032941047290010626, | |
| "grad_norm": 1.6531200408935547, | |
| "learning_rate": 4.87937937937938e-05, | |
| "loss": 2.0848, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.0003307176573163765, | |
| "grad_norm": 0.06442175805568695, | |
| "learning_rate": 4.878878878878879e-05, | |
| "loss": 0.0011, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.0003320248417326468, | |
| "grad_norm": 1.4050544500350952, | |
| "learning_rate": 4.878378378378379e-05, | |
| "loss": 0.779, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.00033333202614891706, | |
| "grad_norm": 1.4547470808029175, | |
| "learning_rate": 4.877877877877878e-05, | |
| "loss": 1.2172, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.0003346392105651873, | |
| "grad_norm": 1.7619500160217285, | |
| "learning_rate": 4.8773773773773776e-05, | |
| "loss": 2.2073, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.0003359463949814576, | |
| "grad_norm": 1.3657172918319702, | |
| "learning_rate": 4.876876876876877e-05, | |
| "loss": 1.0345, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.00033725357939772785, | |
| "grad_norm": 1.9810845851898193, | |
| "learning_rate": 4.876376376376376e-05, | |
| "loss": 2.0455, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.0003385607638139981, | |
| "grad_norm": 1.3832188844680786, | |
| "learning_rate": 4.875875875875876e-05, | |
| "loss": 1.9951, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.0003398679482302684, | |
| "grad_norm": 1.5447360277175903, | |
| "learning_rate": 4.8753753753753754e-05, | |
| "loss": 2.2018, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.00034117513264653865, | |
| "grad_norm": 1.6113301515579224, | |
| "learning_rate": 4.8748748748748754e-05, | |
| "loss": 1.9734, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.0003424823170628089, | |
| "grad_norm": 1.8055421113967896, | |
| "learning_rate": 4.874374374374375e-05, | |
| "loss": 2.1845, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.0003437895014790792, | |
| "grad_norm": 1.568477749824524, | |
| "learning_rate": 4.873873873873874e-05, | |
| "loss": 1.8695, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.00034509668589534945, | |
| "grad_norm": 1.6943508386611938, | |
| "learning_rate": 4.873373373373374e-05, | |
| "loss": 1.8148, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.0003464038703116197, | |
| "grad_norm": 1.0157723426818848, | |
| "learning_rate": 4.8728728728728726e-05, | |
| "loss": 0.6497, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.00034771105472789, | |
| "grad_norm": 1.506629228591919, | |
| "learning_rate": 4.8723723723723726e-05, | |
| "loss": 2.0373, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.00034901823914416024, | |
| "grad_norm": 1.9490463733673096, | |
| "learning_rate": 4.871871871871872e-05, | |
| "loss": 1.37, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.0003503254235604305, | |
| "grad_norm": 1.2232673168182373, | |
| "learning_rate": 4.871371371371372e-05, | |
| "loss": 0.8621, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.0003516326079767007, | |
| "grad_norm": 1.1460094451904297, | |
| "learning_rate": 4.870870870870871e-05, | |
| "loss": 0.6316, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.000352939792392971, | |
| "grad_norm": 1.7706618309020996, | |
| "learning_rate": 4.8703703703703704e-05, | |
| "loss": 1.8774, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.00035424697680924125, | |
| "grad_norm": 1.5327508449554443, | |
| "learning_rate": 4.8698698698698704e-05, | |
| "loss": 1.669, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.0003555541612255115, | |
| "grad_norm": 2.183750629425049, | |
| "learning_rate": 4.86936936936937e-05, | |
| "loss": 1.7982, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.0003568613456417818, | |
| "grad_norm": 1.3268592357635498, | |
| "learning_rate": 4.868868868868869e-05, | |
| "loss": 1.6348, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.00035816853005805205, | |
| "grad_norm": 1.3289145231246948, | |
| "learning_rate": 4.868368368368368e-05, | |
| "loss": 1.0636, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.0003594757144743223, | |
| "grad_norm": 1.543196201324463, | |
| "learning_rate": 4.867867867867868e-05, | |
| "loss": 1.6589, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.0003607828988905926, | |
| "grad_norm": 1.8786145448684692, | |
| "learning_rate": 4.8673673673673676e-05, | |
| "loss": 2.2636, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.00036209008330686284, | |
| "grad_norm": 1.9315189123153687, | |
| "learning_rate": 4.866866866866867e-05, | |
| "loss": 1.4786, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.0003633972677231331, | |
| "grad_norm": 3.0136470794677734, | |
| "learning_rate": 4.866366366366367e-05, | |
| "loss": 1.1689, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.0003647044521394034, | |
| "grad_norm": 1.5813599824905396, | |
| "learning_rate": 4.865865865865866e-05, | |
| "loss": 1.7059, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.00036601163655567364, | |
| "grad_norm": 1.8223422765731812, | |
| "learning_rate": 4.865365365365366e-05, | |
| "loss": 1.2049, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.0003673188209719439, | |
| "grad_norm": 1.1571829319000244, | |
| "learning_rate": 4.8648648648648654e-05, | |
| "loss": 0.9757, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.00036862600538821417, | |
| "grad_norm": 2.1226696968078613, | |
| "learning_rate": 4.864364364364364e-05, | |
| "loss": 2.0413, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.00036993318980448444, | |
| "grad_norm": 1.8020603656768799, | |
| "learning_rate": 4.863863863863864e-05, | |
| "loss": 1.5693, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.0003712403742207547, | |
| "grad_norm": 1.799559235572815, | |
| "learning_rate": 4.863363363363363e-05, | |
| "loss": 2.1869, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.00037254755863702497, | |
| "grad_norm": 2.161092519760132, | |
| "learning_rate": 4.862862862862863e-05, | |
| "loss": 1.903, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.00037385474305329523, | |
| "grad_norm": 1.8967934846878052, | |
| "learning_rate": 4.8623623623623626e-05, | |
| "loss": 1.9437, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.0003751619274695655, | |
| "grad_norm": 1.6356936693191528, | |
| "learning_rate": 4.861861861861862e-05, | |
| "loss": 1.8212, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.00037646911188583576, | |
| "grad_norm": 2.0186562538146973, | |
| "learning_rate": 4.861361361361362e-05, | |
| "loss": 1.9657, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.00037777629630210603, | |
| "grad_norm": 2.0139613151550293, | |
| "learning_rate": 4.860860860860861e-05, | |
| "loss": 1.3713, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.00037908348071837624, | |
| "grad_norm": 1.9778642654418945, | |
| "learning_rate": 4.8603603603603605e-05, | |
| "loss": 2.1371, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.0003803906651346465, | |
| "grad_norm": 2.013936996459961, | |
| "learning_rate": 4.85985985985986e-05, | |
| "loss": 2.1338, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.00038169784955091677, | |
| "grad_norm": 1.480167031288147, | |
| "learning_rate": 4.85935935935936e-05, | |
| "loss": 1.4518, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.00038300503396718704, | |
| "grad_norm": 1.0584605932235718, | |
| "learning_rate": 4.858858858858859e-05, | |
| "loss": 0.9891, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.0003843122183834573, | |
| "grad_norm": 1.5172793865203857, | |
| "learning_rate": 4.858358358358358e-05, | |
| "loss": 1.1713, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.00038561940279972757, | |
| "grad_norm": 1.6944522857666016, | |
| "learning_rate": 4.857857857857858e-05, | |
| "loss": 2.72, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.00038692658721599783, | |
| "grad_norm": 1.6618627309799194, | |
| "learning_rate": 4.8573573573573576e-05, | |
| "loss": 1.3664, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.0003882337716322681, | |
| "grad_norm": 1.5239827632904053, | |
| "learning_rate": 4.8568568568568576e-05, | |
| "loss": 1.8734, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.00038954095604853837, | |
| "grad_norm": 1.4096382856369019, | |
| "learning_rate": 4.856356356356356e-05, | |
| "loss": 1.4521, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.00039084814046480863, | |
| "grad_norm": 1.4097378253936768, | |
| "learning_rate": 4.855855855855856e-05, | |
| "loss": 1.4607, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.0003921553248810789, | |
| "grad_norm": 1.5461453199386597, | |
| "learning_rate": 4.8553553553553555e-05, | |
| "loss": 1.744, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.00039346250929734916, | |
| "grad_norm": 1.9085536003112793, | |
| "learning_rate": 4.854854854854855e-05, | |
| "loss": 2.6346, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.0003947696937136194, | |
| "grad_norm": 1.5974091291427612, | |
| "learning_rate": 4.854354354354355e-05, | |
| "loss": 2.779, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.0003960768781298897, | |
| "grad_norm": 1.4183374643325806, | |
| "learning_rate": 4.853853853853854e-05, | |
| "loss": 1.1477, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.00039738406254615996, | |
| "grad_norm": 1.7596776485443115, | |
| "learning_rate": 4.853353353353354e-05, | |
| "loss": 1.3722, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.0003986912469624302, | |
| "grad_norm": 2.0469233989715576, | |
| "learning_rate": 4.852852852852853e-05, | |
| "loss": 2.532, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.0003999984313787005, | |
| "grad_norm": 1.7058496475219727, | |
| "learning_rate": 4.8523523523523526e-05, | |
| "loss": 1.8808, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.00040130561579497075, | |
| "grad_norm": 2.928122043609619, | |
| "learning_rate": 4.851851851851852e-05, | |
| "loss": 1.4133, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.000402612800211241, | |
| "grad_norm": 2.3317997455596924, | |
| "learning_rate": 4.851351351351351e-05, | |
| "loss": 1.9976, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.0004039199846275113, | |
| "grad_norm": 1.520419716835022, | |
| "learning_rate": 4.850850850850851e-05, | |
| "loss": 1.7102, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.00040522716904378155, | |
| "grad_norm": 1.5943193435668945, | |
| "learning_rate": 4.8503503503503505e-05, | |
| "loss": 1.9234, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.0004065343534600518, | |
| "grad_norm": 0.8694736361503601, | |
| "learning_rate": 4.8498498498498504e-05, | |
| "loss": 0.4603, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.00040784153787632203, | |
| "grad_norm": 1.2508904933929443, | |
| "learning_rate": 4.84934934934935e-05, | |
| "loss": 0.7948, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.0004091487222925923, | |
| "grad_norm": 2.01047420501709, | |
| "learning_rate": 4.848848848848849e-05, | |
| "loss": 1.8304, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.00041045590670886256, | |
| "grad_norm": 2.1941375732421875, | |
| "learning_rate": 4.848348348348349e-05, | |
| "loss": 1.7869, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.0004117630911251328, | |
| "grad_norm": 1.6790601015090942, | |
| "learning_rate": 4.8478478478478476e-05, | |
| "loss": 1.2441, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.0004130702755414031, | |
| "grad_norm": 1.6825275421142578, | |
| "learning_rate": 4.8473473473473476e-05, | |
| "loss": 1.7431, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.00041437745995767336, | |
| "grad_norm": 1.328930377960205, | |
| "learning_rate": 4.846846846846847e-05, | |
| "loss": 1.3589, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.0004156846443739436, | |
| "grad_norm": 1.519547939300537, | |
| "learning_rate": 4.846346346346346e-05, | |
| "loss": 0.5922, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.0004169918287902139, | |
| "grad_norm": 1.8517762422561646, | |
| "learning_rate": 4.845845845845846e-05, | |
| "loss": 2.0995, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.00041829901320648415, | |
| "grad_norm": 1.561224341392517, | |
| "learning_rate": 4.8453453453453455e-05, | |
| "loss": 2.1158, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.0004196061976227544, | |
| "grad_norm": 1.9378067255020142, | |
| "learning_rate": 4.8448448448448455e-05, | |
| "loss": 2.2775, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.0004209133820390247, | |
| "grad_norm": 1.6161144971847534, | |
| "learning_rate": 4.844344344344345e-05, | |
| "loss": 2.2261, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.00042222056645529495, | |
| "grad_norm": 0.9984459280967712, | |
| "learning_rate": 4.843843843843844e-05, | |
| "loss": 0.8248, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.0004235277508715652, | |
| "grad_norm": 1.2390950918197632, | |
| "learning_rate": 4.8433433433433433e-05, | |
| "loss": 0.8056, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.0004248349352878355, | |
| "grad_norm": 1.8470146656036377, | |
| "learning_rate": 4.8428428428428426e-05, | |
| "loss": 1.8804, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.00042614211970410575, | |
| "grad_norm": 1.5264325141906738, | |
| "learning_rate": 4.8423423423423426e-05, | |
| "loss": 1.4915, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.000427449304120376, | |
| "grad_norm": 1.4650579690933228, | |
| "learning_rate": 4.841841841841842e-05, | |
| "loss": 1.9906, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.0004287564885366463, | |
| "grad_norm": 1.6912672519683838, | |
| "learning_rate": 4.841341341341342e-05, | |
| "loss": 2.5512, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.00043006367295291654, | |
| "grad_norm": 1.419115424156189, | |
| "learning_rate": 4.840840840840841e-05, | |
| "loss": 1.254, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.0004313708573691868, | |
| "grad_norm": 1.3889597654342651, | |
| "learning_rate": 4.8403403403403405e-05, | |
| "loss": 1.5753, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.00043267804178545707, | |
| "grad_norm": 1.9419002532958984, | |
| "learning_rate": 4.83983983983984e-05, | |
| "loss": 1.363, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.00043398522620172734, | |
| "grad_norm": 2.337916374206543, | |
| "learning_rate": 4.839339339339339e-05, | |
| "loss": 1.8217, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.00043529241061799755, | |
| "grad_norm": 1.259263277053833, | |
| "learning_rate": 4.838838838838839e-05, | |
| "loss": 1.4361, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.0004365995950342678, | |
| "grad_norm": 1.728857159614563, | |
| "learning_rate": 4.8383383383383384e-05, | |
| "loss": 1.8216, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.0004379067794505381, | |
| "grad_norm": 1.126977562904358, | |
| "learning_rate": 4.837837837837838e-05, | |
| "loss": 0.6098, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.00043921396386680835, | |
| "grad_norm": 1.5760388374328613, | |
| "learning_rate": 4.8373373373373376e-05, | |
| "loss": 1.9721, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0004405211482830786, | |
| "grad_norm": 1.6654877662658691, | |
| "learning_rate": 4.836836836836837e-05, | |
| "loss": 2.0176, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.0004418283326993489, | |
| "grad_norm": 1.5803656578063965, | |
| "learning_rate": 4.836336336336337e-05, | |
| "loss": 1.6791, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.00044313551711561914, | |
| "grad_norm": 1.6455450057983398, | |
| "learning_rate": 4.835835835835836e-05, | |
| "loss": 2.1326, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.0004444427015318894, | |
| "grad_norm": 1.302128553390503, | |
| "learning_rate": 4.8353353353353355e-05, | |
| "loss": 1.3892, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.0004457498859481597, | |
| "grad_norm": 1.2957465648651123, | |
| "learning_rate": 4.834834834834835e-05, | |
| "loss": 1.2624, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.00044705707036442994, | |
| "grad_norm": 1.56611168384552, | |
| "learning_rate": 4.834334334334335e-05, | |
| "loss": 1.2925, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.0004483642547807002, | |
| "grad_norm": 1.5972720384597778, | |
| "learning_rate": 4.833833833833834e-05, | |
| "loss": 2.2345, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.00044967143919697047, | |
| "grad_norm": 1.6864628791809082, | |
| "learning_rate": 4.8333333333333334e-05, | |
| "loss": 2.1596, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.00045097862361324074, | |
| "grad_norm": 1.3631922006607056, | |
| "learning_rate": 4.832832832832833e-05, | |
| "loss": 0.8551, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.000452285808029511, | |
| "grad_norm": 1.2245639562606812, | |
| "learning_rate": 4.8323323323323326e-05, | |
| "loss": 1.613, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.00045359299244578127, | |
| "grad_norm": 1.5226577520370483, | |
| "learning_rate": 4.8318318318318326e-05, | |
| "loss": 1.3359, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.00045490017686205153, | |
| "grad_norm": 1.371559739112854, | |
| "learning_rate": 4.831331331331331e-05, | |
| "loss": 2.1276, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.0004562073612783218, | |
| "grad_norm": 1.8305604457855225, | |
| "learning_rate": 4.830830830830831e-05, | |
| "loss": 2.6008, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.00045751454569459206, | |
| "grad_norm": 1.6536656618118286, | |
| "learning_rate": 4.8303303303303305e-05, | |
| "loss": 1.9033, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.00045882173011086233, | |
| "grad_norm": 1.682886004447937, | |
| "learning_rate": 4.82982982982983e-05, | |
| "loss": 1.7702, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.0004601289145271326, | |
| "grad_norm": 1.8235169649124146, | |
| "learning_rate": 4.82932932932933e-05, | |
| "loss": 2.0686, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.00046143609894340286, | |
| "grad_norm": 1.3382108211517334, | |
| "learning_rate": 4.828828828828829e-05, | |
| "loss": 1.3496, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.0004627432833596731, | |
| "grad_norm": 1.3107976913452148, | |
| "learning_rate": 4.828328328328329e-05, | |
| "loss": 0.7386, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.00046405046777594334, | |
| "grad_norm": 1.5658841133117676, | |
| "learning_rate": 4.8278278278278283e-05, | |
| "loss": 1.1042, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.0004653576521922136, | |
| "grad_norm": 1.51932692527771, | |
| "learning_rate": 4.827327327327327e-05, | |
| "loss": 0.7015, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.00046666483660848387, | |
| "grad_norm": 1.8380029201507568, | |
| "learning_rate": 4.826826826826827e-05, | |
| "loss": 1.8174, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.00046797202102475413, | |
| "grad_norm": 1.3767679929733276, | |
| "learning_rate": 4.826326326326326e-05, | |
| "loss": 1.3915, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.0004692792054410244, | |
| "grad_norm": 1.7605303525924683, | |
| "learning_rate": 4.825825825825826e-05, | |
| "loss": 0.9956, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.00047058638985729466, | |
| "grad_norm": 1.5122482776641846, | |
| "learning_rate": 4.8253253253253255e-05, | |
| "loss": 1.2155, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.00047189357427356493, | |
| "grad_norm": 1.5101649761199951, | |
| "learning_rate": 4.824824824824825e-05, | |
| "loss": 1.1057, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.0004732007586898352, | |
| "grad_norm": 1.0014289617538452, | |
| "learning_rate": 4.824324324324325e-05, | |
| "loss": 0.6814, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.00047450794310610546, | |
| "grad_norm": 1.7926838397979736, | |
| "learning_rate": 4.823823823823824e-05, | |
| "loss": 2.0843, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.0004758151275223757, | |
| "grad_norm": 1.632832407951355, | |
| "learning_rate": 4.823323323323324e-05, | |
| "loss": 2.0709, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.000477122311938646, | |
| "grad_norm": 1.414014458656311, | |
| "learning_rate": 4.822822822822823e-05, | |
| "loss": 1.8019, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.00047842949635491626, | |
| "grad_norm": 1.4348331689834595, | |
| "learning_rate": 4.8223223223223226e-05, | |
| "loss": 1.7161, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.0004797366807711865, | |
| "grad_norm": 1.537306547164917, | |
| "learning_rate": 4.821821821821822e-05, | |
| "loss": 0.8713, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.0004810438651874568, | |
| "grad_norm": 1.8898135423660278, | |
| "learning_rate": 4.821321321321321e-05, | |
| "loss": 2.1803, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.00048235104960372705, | |
| "grad_norm": 1.8930268287658691, | |
| "learning_rate": 4.820820820820821e-05, | |
| "loss": 1.5469, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.0004836582340199973, | |
| "grad_norm": 1.4677132368087769, | |
| "learning_rate": 4.8203203203203205e-05, | |
| "loss": 0.9854, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.0004849654184362676, | |
| "grad_norm": 1.2428200244903564, | |
| "learning_rate": 4.8198198198198205e-05, | |
| "loss": 1.503, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.00048627260285253785, | |
| "grad_norm": 1.542529582977295, | |
| "learning_rate": 4.81931931931932e-05, | |
| "loss": 1.7597, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.0004875797872688081, | |
| "grad_norm": 1.6563135385513306, | |
| "learning_rate": 4.818818818818819e-05, | |
| "loss": 1.2799, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.0004888869716850783, | |
| "grad_norm": 1.8432083129882812, | |
| "learning_rate": 4.8183183183183184e-05, | |
| "loss": 2.2825, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.0004901941561013486, | |
| "grad_norm": 1.8847230672836304, | |
| "learning_rate": 4.817817817817818e-05, | |
| "loss": 1.7107, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.0004915013405176189, | |
| "grad_norm": 1.4317970275878906, | |
| "learning_rate": 4.8173173173173177e-05, | |
| "loss": 1.2853, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.0004928085249338892, | |
| "grad_norm": 1.9522016048431396, | |
| "learning_rate": 4.816816816816817e-05, | |
| "loss": 2.0247, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.0004941157093501594, | |
| "grad_norm": 1.979780912399292, | |
| "learning_rate": 4.816316316316317e-05, | |
| "loss": 2.8973, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.0004954228937664297, | |
| "grad_norm": 1.2153147459030151, | |
| "learning_rate": 4.815815815815816e-05, | |
| "loss": 0.918, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.0004967300781826999, | |
| "grad_norm": 1.7080423831939697, | |
| "learning_rate": 4.8153153153153155e-05, | |
| "loss": 2.3176, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.0004980372625989702, | |
| "grad_norm": 1.4782633781433105, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 1.4934, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.0004993444470152405, | |
| "grad_norm": 1.3005950450897217, | |
| "learning_rate": 4.814314314314314e-05, | |
| "loss": 0.9541, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.0005006516314315108, | |
| "grad_norm": 1.6197017431259155, | |
| "learning_rate": 4.813813813813814e-05, | |
| "loss": 1.5796, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.000501958815847781, | |
| "grad_norm": 1.5193967819213867, | |
| "learning_rate": 4.8133133133133134e-05, | |
| "loss": 1.5869, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.0005032660002640513, | |
| "grad_norm": 2.8697612285614014, | |
| "learning_rate": 4.8128128128128134e-05, | |
| "loss": 1.4886, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.0005045731846803215, | |
| "grad_norm": 1.835699200630188, | |
| "learning_rate": 4.812312312312313e-05, | |
| "loss": 1.6965, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.0005058803690965917, | |
| "grad_norm": 1.51179039478302, | |
| "learning_rate": 4.811811811811812e-05, | |
| "loss": 1.5248, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.000507187553512862, | |
| "grad_norm": 1.3816337585449219, | |
| "learning_rate": 4.811311311311312e-05, | |
| "loss": 1.8567, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.0005084947379291323, | |
| "grad_norm": 1.3670501708984375, | |
| "learning_rate": 4.810810810810811e-05, | |
| "loss": 0.9172, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.0005098019223454026, | |
| "grad_norm": 1.5191140174865723, | |
| "learning_rate": 4.8103103103103105e-05, | |
| "loss": 2.2351, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.0005111091067616728, | |
| "grad_norm": 1.9390945434570312, | |
| "learning_rate": 4.80980980980981e-05, | |
| "loss": 1.5947, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.0005124162911779431, | |
| "grad_norm": 1.0936216115951538, | |
| "learning_rate": 4.809309309309309e-05, | |
| "loss": 0.8947, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.0005137234755942133, | |
| "grad_norm": 3.829148530960083, | |
| "learning_rate": 4.808808808808809e-05, | |
| "loss": 2.7547, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.0005150306600104836, | |
| "grad_norm": 2.89119815826416, | |
| "learning_rate": 4.8083083083083084e-05, | |
| "loss": 1.7366, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.0005163378444267538, | |
| "grad_norm": 1.6022891998291016, | |
| "learning_rate": 4.8078078078078084e-05, | |
| "loss": 1.786, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.0005176450288430242, | |
| "grad_norm": 1.2842696905136108, | |
| "learning_rate": 4.807307307307308e-05, | |
| "loss": 0.2522, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.0005189522132592944, | |
| "grad_norm": 1.976442813873291, | |
| "learning_rate": 4.806806806806807e-05, | |
| "loss": 1.9734, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.0005202593976755647, | |
| "grad_norm": 1.5915790796279907, | |
| "learning_rate": 4.806306306306306e-05, | |
| "loss": 1.9909, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.0005215665820918349, | |
| "grad_norm": 1.658768892288208, | |
| "learning_rate": 4.8058058058058056e-05, | |
| "loss": 1.6076, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.0005228737665081052, | |
| "grad_norm": 1.4726157188415527, | |
| "learning_rate": 4.8053053053053055e-05, | |
| "loss": 2.0277, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0005241809509243754, | |
| "grad_norm": 2.232145309448242, | |
| "learning_rate": 4.804804804804805e-05, | |
| "loss": 1.7958, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.0005254881353406458, | |
| "grad_norm": 1.6406078338623047, | |
| "learning_rate": 4.804304304304305e-05, | |
| "loss": 1.7463, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.000526795319756916, | |
| "grad_norm": 1.1940925121307373, | |
| "learning_rate": 4.803803803803804e-05, | |
| "loss": 0.7934, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.0005281025041731863, | |
| "grad_norm": 1.7259567975997925, | |
| "learning_rate": 4.8033033033033034e-05, | |
| "loss": 0.9874, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.0005294096885894565, | |
| "grad_norm": 1.845804214477539, | |
| "learning_rate": 4.8028028028028034e-05, | |
| "loss": 0.4897, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.0005307168730057268, | |
| "grad_norm": 1.6611266136169434, | |
| "learning_rate": 4.802302302302302e-05, | |
| "loss": 1.7425, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.000532024057421997, | |
| "grad_norm": 1.487190842628479, | |
| "learning_rate": 4.801801801801802e-05, | |
| "loss": 2.0357, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.0005333312418382674, | |
| "grad_norm": 0.8890923261642456, | |
| "learning_rate": 4.801301301301301e-05, | |
| "loss": 0.5583, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.0005346384262545376, | |
| "grad_norm": 1.4603744745254517, | |
| "learning_rate": 4.800800800800801e-05, | |
| "loss": 1.6225, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.0005359456106708078, | |
| "grad_norm": 1.9771097898483276, | |
| "learning_rate": 4.8003003003003005e-05, | |
| "loss": 1.7477, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.0005372527950870781, | |
| "grad_norm": 1.5973646640777588, | |
| "learning_rate": 4.7997997997998e-05, | |
| "loss": 1.5643, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.0005385599795033483, | |
| "grad_norm": 1.9672987461090088, | |
| "learning_rate": 4.7992992992993e-05, | |
| "loss": 2.1034, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.0005398671639196186, | |
| "grad_norm": 1.4598102569580078, | |
| "learning_rate": 4.798798798798799e-05, | |
| "loss": 1.9834, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.0005411743483358888, | |
| "grad_norm": 1.8543391227722168, | |
| "learning_rate": 4.798298298298299e-05, | |
| "loss": 1.8818, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.0005424815327521592, | |
| "grad_norm": 1.3450524806976318, | |
| "learning_rate": 4.797797797797798e-05, | |
| "loss": 1.0332, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.0005437887171684294, | |
| "grad_norm": 1.323103904724121, | |
| "learning_rate": 4.797297297297298e-05, | |
| "loss": 1.4493, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.0005450959015846997, | |
| "grad_norm": 3.834731101989746, | |
| "learning_rate": 4.796796796796797e-05, | |
| "loss": 1.7237, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.0005464030860009699, | |
| "grad_norm": 1.6012156009674072, | |
| "learning_rate": 4.796296296296296e-05, | |
| "loss": 1.7705, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.0005477102704172402, | |
| "grad_norm": 1.6005125045776367, | |
| "learning_rate": 4.795795795795796e-05, | |
| "loss": 1.2692, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.0005490174548335104, | |
| "grad_norm": 0.04926226660609245, | |
| "learning_rate": 4.7952952952952956e-05, | |
| "loss": 0.001, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.0005503246392497808, | |
| "grad_norm": 1.9261682033538818, | |
| "learning_rate": 4.7947947947947955e-05, | |
| "loss": 1.3089, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.000551631823666051, | |
| "grad_norm": 1.4734810590744019, | |
| "learning_rate": 4.794294294294295e-05, | |
| "loss": 1.9636, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.0005529390080823213, | |
| "grad_norm": 1.5746914148330688, | |
| "learning_rate": 4.793793793793794e-05, | |
| "loss": 1.481, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.0005542461924985915, | |
| "grad_norm": 1.6360576152801514, | |
| "learning_rate": 4.7932932932932934e-05, | |
| "loss": 2.0989, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.0005555533769148618, | |
| "grad_norm": 1.9103704690933228, | |
| "learning_rate": 4.792792792792793e-05, | |
| "loss": 2.5168, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.000556860561331132, | |
| "grad_norm": 1.6170457601547241, | |
| "learning_rate": 4.792292292292293e-05, | |
| "loss": 2.1027, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.0005581677457474023, | |
| "grad_norm": 1.6519997119903564, | |
| "learning_rate": 4.791791791791792e-05, | |
| "loss": 1.8065, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.0005594749301636726, | |
| "grad_norm": 2.1365749835968018, | |
| "learning_rate": 4.791291291291291e-05, | |
| "loss": 2.1404, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.0005607821145799429, | |
| "grad_norm": 1.484369158744812, | |
| "learning_rate": 4.790790790790791e-05, | |
| "loss": 1.6508, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.0005620892989962131, | |
| "grad_norm": 2.2153050899505615, | |
| "learning_rate": 4.7902902902902906e-05, | |
| "loss": 1.9617, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.0005633964834124833, | |
| "grad_norm": 1.4849168062210083, | |
| "learning_rate": 4.78978978978979e-05, | |
| "loss": 1.6652, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.0005647036678287536, | |
| "grad_norm": 1.4296141862869263, | |
| "learning_rate": 4.789289289289289e-05, | |
| "loss": 1.2995, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.0005660108522450238, | |
| "grad_norm": 1.6376841068267822, | |
| "learning_rate": 4.788788788788789e-05, | |
| "loss": 2.2176, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.0005673180366612942, | |
| "grad_norm": 1.422378420829773, | |
| "learning_rate": 4.7882882882882884e-05, | |
| "loss": 1.4486, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.0005686252210775644, | |
| "grad_norm": 1.5652416944503784, | |
| "learning_rate": 4.787787787787788e-05, | |
| "loss": 1.8622, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.0005699324054938347, | |
| "grad_norm": 1.2181988954544067, | |
| "learning_rate": 4.787287287287288e-05, | |
| "loss": 0.7264, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.0005712395899101049, | |
| "grad_norm": 1.4485244750976562, | |
| "learning_rate": 4.786786786786787e-05, | |
| "loss": 1.0068, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.0005725467743263752, | |
| "grad_norm": 1.6365692615509033, | |
| "learning_rate": 4.786286286286287e-05, | |
| "loss": 1.8623, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.0005738539587426454, | |
| "grad_norm": 1.362836241722107, | |
| "learning_rate": 4.785785785785786e-05, | |
| "loss": 1.2694, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.0005751611431589157, | |
| "grad_norm": 1.5164575576782227, | |
| "learning_rate": 4.7852852852852856e-05, | |
| "loss": 1.9222, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.000576468327575186, | |
| "grad_norm": 1.7084901332855225, | |
| "learning_rate": 4.784784784784785e-05, | |
| "loss": 1.8659, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.0005777755119914563, | |
| "grad_norm": 2.0331461429595947, | |
| "learning_rate": 4.784284284284284e-05, | |
| "loss": 1.4633, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.0005790826964077265, | |
| "grad_norm": 1.529123306274414, | |
| "learning_rate": 4.783783783783784e-05, | |
| "loss": 1.4277, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.0005803898808239968, | |
| "grad_norm": 1.4111024141311646, | |
| "learning_rate": 4.7832832832832834e-05, | |
| "loss": 1.7261, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.000581697065240267, | |
| "grad_norm": 0.98598313331604, | |
| "learning_rate": 4.7827827827827834e-05, | |
| "loss": 0.4229, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.0005830042496565373, | |
| "grad_norm": 2.0867719650268555, | |
| "learning_rate": 4.782282282282283e-05, | |
| "loss": 1.9944, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.0005843114340728075, | |
| "grad_norm": 1.6054155826568604, | |
| "learning_rate": 4.781781781781782e-05, | |
| "loss": 2.2063, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.0005856186184890779, | |
| "grad_norm": 1.374302864074707, | |
| "learning_rate": 4.781281281281281e-05, | |
| "loss": 1.4218, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.0005869258029053481, | |
| "grad_norm": 1.8737014532089233, | |
| "learning_rate": 4.7807807807807806e-05, | |
| "loss": 2.0067, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.0005882329873216184, | |
| "grad_norm": 1.221209168434143, | |
| "learning_rate": 4.7802802802802806e-05, | |
| "loss": 1.1187, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0005895401717378886, | |
| "grad_norm": 2.31976580619812, | |
| "learning_rate": 4.77977977977978e-05, | |
| "loss": 2.244, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.0005908473561541588, | |
| "grad_norm": 2.0758728981018066, | |
| "learning_rate": 4.77927927927928e-05, | |
| "loss": 2.37, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.0005921545405704291, | |
| "grad_norm": 1.5117443799972534, | |
| "learning_rate": 4.778778778778779e-05, | |
| "loss": 1.4368, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.0005934617249866994, | |
| "grad_norm": 2.0739381313323975, | |
| "learning_rate": 4.7782782782782784e-05, | |
| "loss": 2.3445, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.0005947689094029697, | |
| "grad_norm": 1.6032236814498901, | |
| "learning_rate": 4.7777777777777784e-05, | |
| "loss": 1.3034, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.0005960760938192399, | |
| "grad_norm": 1.8953197002410889, | |
| "learning_rate": 4.777277277277277e-05, | |
| "loss": 2.4014, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.0005973832782355102, | |
| "grad_norm": 1.1892727613449097, | |
| "learning_rate": 4.776776776776777e-05, | |
| "loss": 1.0365, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.0005986904626517804, | |
| "grad_norm": 1.3845131397247314, | |
| "learning_rate": 4.776276276276276e-05, | |
| "loss": 1.4982, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.0005999976470680507, | |
| "grad_norm": 2.0147531032562256, | |
| "learning_rate": 4.775775775775776e-05, | |
| "loss": 1.0432, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.0006013048314843209, | |
| "grad_norm": 1.3901069164276123, | |
| "learning_rate": 4.7752752752752756e-05, | |
| "loss": 1.1874, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.0006026120159005913, | |
| "grad_norm": 1.7510122060775757, | |
| "learning_rate": 4.774774774774775e-05, | |
| "loss": 1.7532, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.0006039192003168615, | |
| "grad_norm": 1.4255083799362183, | |
| "learning_rate": 4.774274274274275e-05, | |
| "loss": 1.2638, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.0006052263847331318, | |
| "grad_norm": 1.3789587020874023, | |
| "learning_rate": 4.773773773773774e-05, | |
| "loss": 1.2885, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.000606533569149402, | |
| "grad_norm": 1.8223754167556763, | |
| "learning_rate": 4.7732732732732734e-05, | |
| "loss": 1.846, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.0006078407535656723, | |
| "grad_norm": 1.9569519758224487, | |
| "learning_rate": 4.772772772772773e-05, | |
| "loss": 2.0224, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.0006091479379819425, | |
| "grad_norm": 1.589490532875061, | |
| "learning_rate": 4.772272272272272e-05, | |
| "loss": 1.4883, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.0006104551223982129, | |
| "grad_norm": 1.4150370359420776, | |
| "learning_rate": 4.771771771771772e-05, | |
| "loss": 1.1537, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.0006117623068144831, | |
| "grad_norm": 1.3750197887420654, | |
| "learning_rate": 4.771271271271271e-05, | |
| "loss": 1.6263, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.0006130694912307534, | |
| "grad_norm": 1.5440406799316406, | |
| "learning_rate": 4.770770770770771e-05, | |
| "loss": 1.6786, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.0006143766756470236, | |
| "grad_norm": 1.7556005716323853, | |
| "learning_rate": 4.7702702702702706e-05, | |
| "loss": 2.0755, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.0006156838600632939, | |
| "grad_norm": 1.2925258874893188, | |
| "learning_rate": 4.76976976976977e-05, | |
| "loss": 0.8731, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.0006169910444795641, | |
| "grad_norm": 1.463951826095581, | |
| "learning_rate": 4.76926926926927e-05, | |
| "loss": 1.7757, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.0006182982288958343, | |
| "grad_norm": 1.1145371198654175, | |
| "learning_rate": 4.7687687687687685e-05, | |
| "loss": 0.3717, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.0006196054133121047, | |
| "grad_norm": 1.656201720237732, | |
| "learning_rate": 4.7682682682682685e-05, | |
| "loss": 1.3816, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.0006209125977283749, | |
| "grad_norm": 1.6448125839233398, | |
| "learning_rate": 4.767767767767768e-05, | |
| "loss": 1.3373, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.0006222197821446452, | |
| "grad_norm": 1.5466721057891846, | |
| "learning_rate": 4.767267267267268e-05, | |
| "loss": 1.757, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.0006235269665609154, | |
| "grad_norm": 1.4172669649124146, | |
| "learning_rate": 4.766766766766767e-05, | |
| "loss": 1.3798, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.0006248341509771857, | |
| "grad_norm": 1.4851760864257812, | |
| "learning_rate": 4.766266266266266e-05, | |
| "loss": 1.5471, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.0006261413353934559, | |
| "grad_norm": 1.2650772333145142, | |
| "learning_rate": 4.765765765765766e-05, | |
| "loss": 0.5324, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.0006274485198097263, | |
| "grad_norm": 1.645494818687439, | |
| "learning_rate": 4.7652652652652656e-05, | |
| "loss": 1.422, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.0006287557042259965, | |
| "grad_norm": 1.4462095499038696, | |
| "learning_rate": 4.764764764764765e-05, | |
| "loss": 0.7855, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.0006300628886422668, | |
| "grad_norm": 1.6467872858047485, | |
| "learning_rate": 4.764264264264264e-05, | |
| "loss": 1.7378, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.000631370073058537, | |
| "grad_norm": 1.4757755994796753, | |
| "learning_rate": 4.763763763763764e-05, | |
| "loss": 1.2541, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.0006326772574748073, | |
| "grad_norm": 1.649613618850708, | |
| "learning_rate": 4.7632632632632635e-05, | |
| "loss": 1.9402, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.0006339844418910775, | |
| "grad_norm": 1.5895005464553833, | |
| "learning_rate": 4.762762762762763e-05, | |
| "loss": 1.6001, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.0006352916263073478, | |
| "grad_norm": 1.9634181261062622, | |
| "learning_rate": 4.762262262262263e-05, | |
| "loss": 1.9354, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.0006365988107236181, | |
| "grad_norm": 1.5781902074813843, | |
| "learning_rate": 4.761761761761762e-05, | |
| "loss": 1.7391, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.0006379059951398884, | |
| "grad_norm": 1.7558811902999878, | |
| "learning_rate": 4.761261261261262e-05, | |
| "loss": 1.3157, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.0006392131795561586, | |
| "grad_norm": 1.4790571928024292, | |
| "learning_rate": 4.7607607607607606e-05, | |
| "loss": 1.2889, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.0006405203639724289, | |
| "grad_norm": 1.9054185152053833, | |
| "learning_rate": 4.7602602602602606e-05, | |
| "loss": 1.9227, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.0006418275483886991, | |
| "grad_norm": 1.3396105766296387, | |
| "learning_rate": 4.75975975975976e-05, | |
| "loss": 0.966, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.0006431347328049694, | |
| "grad_norm": 1.6935501098632812, | |
| "learning_rate": 4.759259259259259e-05, | |
| "loss": 1.3237, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.0006444419172212397, | |
| "grad_norm": 1.58639395236969, | |
| "learning_rate": 4.758758758758759e-05, | |
| "loss": 1.8338, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.0006457491016375099, | |
| "grad_norm": 1.8331503868103027, | |
| "learning_rate": 4.7582582582582585e-05, | |
| "loss": 1.5735, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.0006470562860537802, | |
| "grad_norm": 2.0812811851501465, | |
| "learning_rate": 4.7577577577577584e-05, | |
| "loss": 2.0272, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.0006483634704700504, | |
| "grad_norm": 1.8232816457748413, | |
| "learning_rate": 4.757257257257258e-05, | |
| "loss": 1.52, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.0006496706548863207, | |
| "grad_norm": 1.5954536199569702, | |
| "learning_rate": 4.756756756756757e-05, | |
| "loss": 1.8882, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.0006509778393025909, | |
| "grad_norm": 2.2882657051086426, | |
| "learning_rate": 4.756256256256256e-05, | |
| "loss": 2.3967, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.0006522850237188612, | |
| "grad_norm": 1.4601846933364868, | |
| "learning_rate": 4.7557557557557556e-05, | |
| "loss": 1.7912, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.0006535922081351315, | |
| "grad_norm": 1.9598947763442993, | |
| "learning_rate": 4.7552552552552556e-05, | |
| "loss": 2.0452, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 439137192394752.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |