| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.0, |
| "eval_steps": 500, |
| "global_step": 150, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.02, |
| "grad_norm": 0.009636741336022452, |
| "learning_rate": 5.333333333333334e-06, |
| "loss": 0.1715, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 0.010362687517832486, |
| "learning_rate": 1.0666666666666667e-05, |
| "loss": 0.1824, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.06, |
| "grad_norm": 0.009198210804198782, |
| "learning_rate": 1.6000000000000003e-05, |
| "loss": 0.1754, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 0.009465239071314193, |
| "learning_rate": 2.1333333333333335e-05, |
| "loss": 0.1755, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.1, |
| "grad_norm": 0.009478588702774012, |
| "learning_rate": 2.6666666666666667e-05, |
| "loss": 0.1853, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 0.010099136115960817, |
| "learning_rate": 3.2000000000000005e-05, |
| "loss": 0.1801, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.14, |
| "grad_norm": 0.011613644980684073, |
| "learning_rate": 3.733333333333334e-05, |
| "loss": 0.176, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 0.01197743565023999, |
| "learning_rate": 4.266666666666667e-05, |
| "loss": 0.1777, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.18, |
| "grad_norm": 0.013570201775932841, |
| "learning_rate": 4.8e-05, |
| "loss": 0.1732, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 0.01341025447595714, |
| "learning_rate": 5.333333333333333e-05, |
| "loss": 0.1768, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.22, |
| "grad_norm": 0.013501878402477732, |
| "learning_rate": 5.8666666666666665e-05, |
| "loss": 0.1791, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 0.01394807803763648, |
| "learning_rate": 6.400000000000001e-05, |
| "loss": 0.1628, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.26, |
| "grad_norm": 0.01170610516898677, |
| "learning_rate": 6.933333333333334e-05, |
| "loss": 0.1751, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 0.011098547548776561, |
| "learning_rate": 7.466666666666667e-05, |
| "loss": 0.1563, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.3, |
| "grad_norm": 0.010378027800969287, |
| "learning_rate": 8e-05, |
| "loss": 0.1587, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 0.014591340667016275, |
| "learning_rate": 7.998916964717848e-05, |
| "loss": 0.1567, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.34, |
| "grad_norm": 0.015066581152101485, |
| "learning_rate": 7.9956684453541e-05, |
| "loss": 0.1542, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 0.01935631722733039, |
| "learning_rate": 7.990256201039297e-05, |
| "loss": 0.1395, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.38, |
| "grad_norm": 0.016114690369511887, |
| "learning_rate": 7.982683162599218e-05, |
| "loss": 0.1487, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 0.01521055999956516, |
| "learning_rate": 7.972953430967773e-05, |
| "loss": 0.1362, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.42, |
| "grad_norm": 0.01388400686918616, |
| "learning_rate": 7.961072274966282e-05, |
| "loss": 0.1323, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 0.011803776763410411, |
| "learning_rate": 7.947046128450319e-05, |
| "loss": 0.1281, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.46, |
| "grad_norm": 0.011045288797632963, |
| "learning_rate": 7.930882586825653e-05, |
| "loss": 0.1201, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 0.010599083830245796, |
| "learning_rate": 7.912590402935223e-05, |
| "loss": 0.132, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 0.011103137449568937, |
| "learning_rate": 7.892179482319297e-05, |
| "loss": 0.1141, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 0.011902802340871899, |
| "learning_rate": 7.869660877851456e-05, |
| "loss": 0.1219, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.54, |
| "grad_norm": 0.012451130022635816, |
| "learning_rate": 7.845046783753276e-05, |
| "loss": 0.1008, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 0.011143892428532776, |
| "learning_rate": 7.818350528990929e-05, |
| "loss": 0.0997, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.58, |
| "grad_norm": 0.010372347838618078, |
| "learning_rate": 7.789586570057317e-05, |
| "loss": 0.0957, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 0.009676864383070798, |
| "learning_rate": 7.758770483143634e-05, |
| "loss": 0.0925, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.62, |
| "grad_norm": 0.008672563061085287, |
| "learning_rate": 7.72591895570457e-05, |
| "loss": 0.0973, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 0.00898212793156372, |
| "learning_rate": 7.69104977742177e-05, |
| "loss": 0.0974, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.66, |
| "grad_norm": 0.009109303563397697, |
| "learning_rate": 7.654181830570404e-05, |
| "loss": 0.0998, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 0.00993359401368422, |
| "learning_rate": 7.615335079794083e-05, |
| "loss": 0.086, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.7, |
| "grad_norm": 0.009207584274834493, |
| "learning_rate": 7.57453056129365e-05, |
| "loss": 0.0931, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 0.008893102280048583, |
| "learning_rate": 7.531790371435709e-05, |
| "loss": 0.0926, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.74, |
| "grad_norm": 0.008990839608298584, |
| "learning_rate": 7.48713765478705e-05, |
| "loss": 0.0758, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 0.009328548294862071, |
| "learning_rate": 7.440596591581463e-05, |
| "loss": 0.0827, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.78, |
| "grad_norm": 0.00895683248045566, |
| "learning_rate": 7.392192384625704e-05, |
| "loss": 0.0815, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 0.01562957245912284, |
| "learning_rate": 7.341951245651747e-05, |
| "loss": 0.0822, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.82, |
| "grad_norm": 0.011492930303382679, |
| "learning_rate": 7.28990038112265e-05, |
| "loss": 0.0819, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 0.009720394941020897, |
| "learning_rate": 7.236067977499791e-05, |
| "loss": 0.0711, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.86, |
| "grad_norm": 0.009306557079384492, |
| "learning_rate": 7.180483185979392e-05, |
| "loss": 0.0739, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 0.01001901077596384, |
| "learning_rate": 7.123176106706638e-05, |
| "loss": 0.0691, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.9, |
| "grad_norm": 0.010826427771236737, |
| "learning_rate": 7.064177772475912e-05, |
| "loss": 0.0615, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 0.011747171511109867, |
| "learning_rate": 7.003520131925997e-05, |
| "loss": 0.075, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.94, |
| "grad_norm": 0.009651751755922859, |
| "learning_rate": 6.941236032239316e-05, |
| "loss": 0.0748, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 0.011647724272264882, |
| "learning_rate": 6.877359201354606e-05, |
| "loss": 0.0641, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.98, |
| "grad_norm": 0.010044576414893541, |
| "learning_rate": 6.811924229702648e-05, |
| "loss": 0.0557, |
| "step": 49 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.010963236047344335, |
| "learning_rate": 6.744966551474936e-05, |
| "loss": 0.0745, |
| "step": 50 |
| }, |
| { |
| "epoch": 1.02, |
| "grad_norm": 0.011693408393284645, |
| "learning_rate": 6.676522425435433e-05, |
| "loss": 0.0637, |
| "step": 51 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 0.010041015684255292, |
| "learning_rate": 6.606628915285822e-05, |
| "loss": 0.0635, |
| "step": 52 |
| }, |
| { |
| "epoch": 1.06, |
| "grad_norm": 0.01214521226945124, |
| "learning_rate": 6.53532386959484e-05, |
| "loss": 0.0625, |
| "step": 53 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 0.013412576285306434, |
| "learning_rate": 6.462645901302633e-05, |
| "loss": 0.059, |
| "step": 54 |
| }, |
| { |
| "epoch": 1.1, |
| "grad_norm": 0.012492830990301382, |
| "learning_rate": 6.388634366811146e-05, |
| "loss": 0.0664, |
| "step": 55 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 0.010307881080182013, |
| "learning_rate": 6.313329344671946e-05, |
| "loss": 0.055, |
| "step": 56 |
| }, |
| { |
| "epoch": 1.1400000000000001, |
| "grad_norm": 0.014154063972213174, |
| "learning_rate": 6.236771613882987e-05, |
| "loss": 0.0647, |
| "step": 57 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 0.00758178084492694, |
| "learning_rate": 6.159002631806052e-05, |
| "loss": 0.0538, |
| "step": 58 |
| }, |
| { |
| "epoch": 1.18, |
| "grad_norm": 0.00880854740589445, |
| "learning_rate": 6.0800645117168616e-05, |
| "loss": 0.056, |
| "step": 59 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 0.009023772056842088, |
| "learning_rate": 6.000000000000001e-05, |
| "loss": 0.0759, |
| "step": 60 |
| }, |
| { |
| "epoch": 1.22, |
| "grad_norm": 0.010378314572008647, |
| "learning_rate": 5.918852453000986e-05, |
| "loss": 0.0496, |
| "step": 61 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 0.012072477515037063, |
| "learning_rate": 5.836665813548047e-05, |
| "loss": 0.0621, |
| "step": 62 |
| }, |
| { |
| "epoch": 1.26, |
| "grad_norm": 0.009248961092533802, |
| "learning_rate": 5.75348458715631e-05, |
| "loss": 0.0561, |
| "step": 63 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 0.015091312399039424, |
| "learning_rate": 5.669353817927272e-05, |
| "loss": 0.0557, |
| "step": 64 |
| }, |
| { |
| "epoch": 1.3, |
| "grad_norm": 0.0070923811566205325, |
| "learning_rate": 5.584319064156628e-05, |
| "loss": 0.0615, |
| "step": 65 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 0.007952025290001136, |
| "learning_rate": 5.4984263736636494e-05, |
| "loss": 0.057, |
| "step": 66 |
| }, |
| { |
| "epoch": 1.34, |
| "grad_norm": 0.007939353667071943, |
| "learning_rate": 5.4117222588554756e-05, |
| "loss": 0.0546, |
| "step": 67 |
| }, |
| { |
| "epoch": 1.3599999999999999, |
| "grad_norm": 0.006919385279452628, |
| "learning_rate": 5.324253671539833e-05, |
| "loss": 0.062, |
| "step": 68 |
| }, |
| { |
| "epoch": 1.38, |
| "grad_norm": 0.007089604487267885, |
| "learning_rate": 5.23606797749979e-05, |
| "loss": 0.0598, |
| "step": 69 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 0.008623592934742961, |
| "learning_rate": 5.1472129308443616e-05, |
| "loss": 0.0551, |
| "step": 70 |
| }, |
| { |
| "epoch": 1.42, |
| "grad_norm": 0.008784978928849127, |
| "learning_rate": 5.05773664814881e-05, |
| "loss": 0.061, |
| "step": 71 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 0.006550166798152803, |
| "learning_rate": 4.967687582398671e-05, |
| "loss": 0.0622, |
| "step": 72 |
| }, |
| { |
| "epoch": 1.46, |
| "grad_norm": 0.007052277272565447, |
| "learning_rate": 4.877114496751613e-05, |
| "loss": 0.067, |
| "step": 73 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 0.006097398821603674, |
| "learning_rate": 4.786066438131321e-05, |
| "loss": 0.0614, |
| "step": 74 |
| }, |
| { |
| "epoch": 1.5, |
| "grad_norm": 0.008263078087014087, |
| "learning_rate": 4.694592710667723e-05, |
| "loss": 0.0525, |
| "step": 75 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 0.00715066783054298, |
| "learning_rate": 4.602742848997933e-05, |
| "loss": 0.0643, |
| "step": 76 |
| }, |
| { |
| "epoch": 1.54, |
| "grad_norm": 0.0067154327511224745, |
| "learning_rate": 4.51056659144238e-05, |
| "loss": 0.0588, |
| "step": 77 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 0.00707798902237324, |
| "learning_rate": 4.418113853070614e-05, |
| "loss": 0.0615, |
| "step": 78 |
| }, |
| { |
| "epoch": 1.58, |
| "grad_norm": 0.0067307685819475984, |
| "learning_rate": 4.3254346986714334e-05, |
| "loss": 0.0582, |
| "step": 79 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 0.01381082737610857, |
| "learning_rate": 4.2325793156419035e-05, |
| "loss": 0.0553, |
| "step": 80 |
| }, |
| { |
| "epoch": 1.62, |
| "grad_norm": 0.00623373080295011, |
| "learning_rate": 4.139597986810005e-05, |
| "loss": 0.0604, |
| "step": 81 |
| }, |
| { |
| "epoch": 1.6400000000000001, |
| "grad_norm": 0.00629872507047934, |
| "learning_rate": 4.046541063205589e-05, |
| "loss": 0.0619, |
| "step": 82 |
| }, |
| { |
| "epoch": 1.6600000000000001, |
| "grad_norm": 0.009046273412923134, |
| "learning_rate": 3.953458936794413e-05, |
| "loss": 0.0517, |
| "step": 83 |
| }, |
| { |
| "epoch": 1.6800000000000002, |
| "grad_norm": 0.006483406280602329, |
| "learning_rate": 3.860402013189998e-05, |
| "loss": 0.0585, |
| "step": 84 |
| }, |
| { |
| "epoch": 1.7, |
| "grad_norm": 0.0067897402127587705, |
| "learning_rate": 3.767420684358097e-05, |
| "loss": 0.0543, |
| "step": 85 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 0.00649969427515724, |
| "learning_rate": 3.674565301328568e-05, |
| "loss": 0.0588, |
| "step": 86 |
| }, |
| { |
| "epoch": 1.74, |
| "grad_norm": 0.007060574694883647, |
| "learning_rate": 3.581886146929387e-05, |
| "loss": 0.0633, |
| "step": 87 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 0.006392848064890084, |
| "learning_rate": 3.4894334085576215e-05, |
| "loss": 0.06, |
| "step": 88 |
| }, |
| { |
| "epoch": 1.78, |
| "grad_norm": 0.006618394605316045, |
| "learning_rate": 3.397257151002068e-05, |
| "loss": 0.0528, |
| "step": 89 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 0.006697234159476227, |
| "learning_rate": 3.305407289332279e-05, |
| "loss": 0.0498, |
| "step": 90 |
| }, |
| { |
| "epoch": 1.8199999999999998, |
| "grad_norm": 0.00610530129114645, |
| "learning_rate": 3.213933561868679e-05, |
| "loss": 0.0577, |
| "step": 91 |
| }, |
| { |
| "epoch": 1.8399999999999999, |
| "grad_norm": 0.007418991369044916, |
| "learning_rate": 3.122885503248386e-05, |
| "loss": 0.0587, |
| "step": 92 |
| }, |
| { |
| "epoch": 1.8599999999999999, |
| "grad_norm": 0.0070687500804459385, |
| "learning_rate": 3.0323124176013297e-05, |
| "loss": 0.0602, |
| "step": 93 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 0.006123140049340437, |
| "learning_rate": 2.9422633518511926e-05, |
| "loss": 0.0695, |
| "step": 94 |
| }, |
| { |
| "epoch": 1.9, |
| "grad_norm": 0.006298273561172453, |
| "learning_rate": 2.8527870691556404e-05, |
| "loss": 0.0584, |
| "step": 95 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 0.0067135751792378705, |
| "learning_rate": 2.7639320225002108e-05, |
| "loss": 0.0503, |
| "step": 96 |
| }, |
| { |
| "epoch": 1.94, |
| "grad_norm": 0.006282664432727376, |
| "learning_rate": 2.6757463284601682e-05, |
| "loss": 0.0612, |
| "step": 97 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 0.00614331973491928, |
| "learning_rate": 2.5882777411445254e-05, |
| "loss": 0.0557, |
| "step": 98 |
| }, |
| { |
| "epoch": 1.98, |
| "grad_norm": 0.006513674627767057, |
| "learning_rate": 2.501573626336352e-05, |
| "loss": 0.0621, |
| "step": 99 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.006042169788173783, |
| "learning_rate": 2.4156809358433728e-05, |
| "loss": 0.0554, |
| "step": 100 |
| }, |
| { |
| "epoch": 2.02, |
| "grad_norm": 0.006044407665434084, |
| "learning_rate": 2.330646182072729e-05, |
| "loss": 0.056, |
| "step": 101 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 0.006233588087031301, |
| "learning_rate": 2.24651541284369e-05, |
| "loss": 0.0523, |
| "step": 102 |
| }, |
| { |
| "epoch": 2.06, |
| "grad_norm": 0.006602363346424627, |
| "learning_rate": 2.1633341864519526e-05, |
| "loss": 0.057, |
| "step": 103 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 0.006164413712347329, |
| "learning_rate": 2.0811475469990167e-05, |
| "loss": 0.0642, |
| "step": 104 |
| }, |
| { |
| "epoch": 2.1, |
| "grad_norm": 0.006231871539708628, |
| "learning_rate": 2.0000000000000012e-05, |
| "loss": 0.0571, |
| "step": 105 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 0.006780552445760737, |
| "learning_rate": 1.9199354882831387e-05, |
| "loss": 0.0668, |
| "step": 106 |
| }, |
| { |
| "epoch": 2.14, |
| "grad_norm": 0.006210172704867486, |
| "learning_rate": 1.8409973681939498e-05, |
| "loss": 0.0515, |
| "step": 107 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 0.00635625124587468, |
| "learning_rate": 1.7632283861170135e-05, |
| "loss": 0.0602, |
| "step": 108 |
| }, |
| { |
| "epoch": 2.18, |
| "grad_norm": 0.006099019550879684, |
| "learning_rate": 1.686670655328054e-05, |
| "loss": 0.0567, |
| "step": 109 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 0.006148019770628567, |
| "learning_rate": 1.6113656331888563e-05, |
| "loss": 0.0543, |
| "step": 110 |
| }, |
| { |
| "epoch": 2.22, |
| "grad_norm": 0.006188474555622255, |
| "learning_rate": 1.537354098697367e-05, |
| "loss": 0.0513, |
| "step": 111 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 0.0062800325948182125, |
| "learning_rate": 1.4646761304051587e-05, |
| "loss": 0.0569, |
| "step": 112 |
| }, |
| { |
| "epoch": 2.26, |
| "grad_norm": 0.006650988867652313, |
| "learning_rate": 1.3933710847141795e-05, |
| "loss": 0.0498, |
| "step": 113 |
| }, |
| { |
| "epoch": 2.2800000000000002, |
| "grad_norm": 0.006505825700635924, |
| "learning_rate": 1.3234775745645684e-05, |
| "loss": 0.0559, |
| "step": 114 |
| }, |
| { |
| "epoch": 2.3, |
| "grad_norm": 0.00643511942149351, |
| "learning_rate": 1.2550334485250661e-05, |
| "loss": 0.058, |
| "step": 115 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 0.006703346043408033, |
| "learning_rate": 1.1880757702973531e-05, |
| "loss": 0.0514, |
| "step": 116 |
| }, |
| { |
| "epoch": 2.34, |
| "grad_norm": 0.006614833898072401, |
| "learning_rate": 1.1226407986453963e-05, |
| "loss": 0.048, |
| "step": 117 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 0.008421863689815894, |
| "learning_rate": 1.0587639677606857e-05, |
| "loss": 0.0569, |
| "step": 118 |
| }, |
| { |
| "epoch": 2.38, |
| "grad_norm": 0.00702647634214854, |
| "learning_rate": 9.964798680740033e-06, |
| "loss": 0.0607, |
| "step": 119 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 0.007525672678270883, |
| "learning_rate": 9.358222275240884e-06, |
| "loss": 0.0516, |
| "step": 120 |
| }, |
| { |
| "epoch": 2.42, |
| "grad_norm": 0.0071990855310353355, |
| "learning_rate": 8.768238932933632e-06, |
| "loss": 0.0561, |
| "step": 121 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 0.006676278729431774, |
| "learning_rate": 8.195168140206084e-06, |
| "loss": 0.0508, |
| "step": 122 |
| }, |
| { |
| "epoch": 2.46, |
| "grad_norm": 0.006726756383711495, |
| "learning_rate": 7.639320225002106e-06, |
| "loss": 0.0543, |
| "step": 123 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 0.006478102996491662, |
| "learning_rate": 7.1009961887735075e-06, |
| "loss": 0.0615, |
| "step": 124 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 0.006384672199557418, |
| "learning_rate": 6.58048754348255e-06, |
| "loss": 0.0519, |
| "step": 125 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 0.006417102700792408, |
| "learning_rate": 6.078076153742962e-06, |
| "loss": 0.061, |
| "step": 126 |
| }, |
| { |
| "epoch": 2.54, |
| "grad_norm": 0.007067596970480207, |
| "learning_rate": 5.5940340841853915e-06, |
| "loss": 0.0489, |
| "step": 127 |
| }, |
| { |
| "epoch": 2.56, |
| "grad_norm": 0.0071492930770040404, |
| "learning_rate": 5.128623452129508e-06, |
| "loss": 0.063, |
| "step": 128 |
| }, |
| { |
| "epoch": 2.58, |
| "grad_norm": 0.0070583015238891945, |
| "learning_rate": 4.6820962856429205e-06, |
| "loss": 0.0566, |
| "step": 129 |
| }, |
| { |
| "epoch": 2.6, |
| "grad_norm": 0.006106875871626197, |
| "learning_rate": 4.254694387063514e-06, |
| "loss": 0.057, |
| "step": 130 |
| }, |
| { |
| "epoch": 2.62, |
| "grad_norm": 0.006657825061725054, |
| "learning_rate": 3.846649202059181e-06, |
| "loss": 0.0552, |
| "step": 131 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 0.006550242431690728, |
| "learning_rate": 3.458181694295961e-06, |
| "loss": 0.0537, |
| "step": 132 |
| }, |
| { |
| "epoch": 2.66, |
| "grad_norm": 0.0072149762384190195, |
| "learning_rate": 3.0895022257823083e-06, |
| "loss": 0.046, |
| "step": 133 |
| }, |
| { |
| "epoch": 2.68, |
| "grad_norm": 0.006643776884830298, |
| "learning_rate": 2.7408104429543025e-06, |
| "loss": 0.0588, |
| "step": 134 |
| }, |
| { |
| "epoch": 2.7, |
| "grad_norm": 0.006712189275186943, |
| "learning_rate": 2.4122951685636674e-06, |
| "loss": 0.0653, |
| "step": 135 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "grad_norm": 0.006801932431398963, |
| "learning_rate": 2.104134299426832e-06, |
| "loss": 0.0515, |
| "step": 136 |
| }, |
| { |
| "epoch": 2.74, |
| "grad_norm": 0.0066739185560487704, |
| "learning_rate": 1.8164947100907238e-06, |
| "loss": 0.0599, |
| "step": 137 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 0.006627424538532696, |
| "learning_rate": 1.5495321624672443e-06, |
| "loss": 0.0614, |
| "step": 138 |
| }, |
| { |
| "epoch": 2.7800000000000002, |
| "grad_norm": 0.007231884663539351, |
| "learning_rate": 1.3033912214854482e-06, |
| "loss": 0.046, |
| "step": 139 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 0.007846672987071566, |
| "learning_rate": 1.0782051768070477e-06, |
| "loss": 0.0526, |
| "step": 140 |
| }, |
| { |
| "epoch": 2.82, |
| "grad_norm": 0.006460619883406472, |
| "learning_rate": 8.740959706477725e-07, |
| "loss": 0.0541, |
| "step": 141 |
| }, |
| { |
| "epoch": 2.84, |
| "grad_norm": 0.0065134704439769155, |
| "learning_rate": 6.911741317434706e-07, |
| "loss": 0.0518, |
| "step": 142 |
| }, |
| { |
| "epoch": 2.86, |
| "grad_norm": 0.006813223503825361, |
| "learning_rate": 5.295387154968312e-07, |
| "loss": 0.0495, |
| "step": 143 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 0.0065749686524116105, |
| "learning_rate": 3.8927725033718553e-07, |
| "loss": 0.0545, |
| "step": 144 |
| }, |
| { |
| "epoch": 2.9, |
| "grad_norm": 0.006958602579708186, |
| "learning_rate": 2.704656903222791e-07, |
| "loss": 0.0489, |
| "step": 145 |
| }, |
| { |
| "epoch": 2.92, |
| "grad_norm": 0.00748378605411929, |
| "learning_rate": 1.7316837400782604e-07, |
| "loss": 0.0577, |
| "step": 146 |
| }, |
| { |
| "epoch": 2.94, |
| "grad_norm": 0.006538362949209428, |
| "learning_rate": 9.74379896070321e-08, |
| "loss": 0.0543, |
| "step": 147 |
| }, |
| { |
| "epoch": 2.96, |
| "grad_norm": 0.006585731397873575, |
| "learning_rate": 4.331554645901737e-08, |
| "loss": 0.0585, |
| "step": 148 |
| }, |
| { |
| "epoch": 2.98, |
| "grad_norm": 0.0066164690655662476, |
| "learning_rate": 1.0830352821531442e-08, |
| "loss": 0.0527, |
| "step": 149 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 0.006825834210227582, |
| "learning_rate": 0.0, |
| "loss": 0.0467, |
| "step": 150 |
| }, |
| { |
| "epoch": 3.0, |
| "step": 150, |
| "total_flos": 112968873082880.0, |
| "train_loss": 0.07854529599348704, |
| "train_runtime": 1238.0931, |
| "train_samples_per_second": 0.969, |
| "train_steps_per_second": 0.121 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 150, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 112968873082880.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
