| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.6456390565002743, | |
| "eval_steps": 1000, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013713658804168952, | |
| "grad_norm": 0.9250810742378235, | |
| "learning_rate": 2.4808205470313544e-05, | |
| "loss": 2.5905, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.027427317608337904, | |
| "grad_norm": 0.7566132545471191, | |
| "learning_rate": 2.459973315543696e-05, | |
| "loss": 2.1754, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04114097641250686, | |
| "grad_norm": 0.8571016788482666, | |
| "learning_rate": 2.4391260840560377e-05, | |
| "loss": 2.063, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.05485463521667581, | |
| "grad_norm": 0.734812319278717, | |
| "learning_rate": 2.418278852568379e-05, | |
| "loss": 2.0424, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06856829402084476, | |
| "grad_norm": 0.706753671169281, | |
| "learning_rate": 2.3974316210807207e-05, | |
| "loss": 2.022, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08228195282501372, | |
| "grad_norm": 0.7179118394851685, | |
| "learning_rate": 2.3765843895930624e-05, | |
| "loss": 2.0341, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09599561162918267, | |
| "grad_norm": 0.7341741323471069, | |
| "learning_rate": 2.3557371581054037e-05, | |
| "loss": 2.0339, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.10970927043335162, | |
| "grad_norm": 0.6983020901679993, | |
| "learning_rate": 2.334889926617745e-05, | |
| "loss": 1.9717, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12342292923752057, | |
| "grad_norm": 0.7121065855026245, | |
| "learning_rate": 2.3140426951300867e-05, | |
| "loss": 1.996, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.13713658804168952, | |
| "grad_norm": 0.7687849402427673, | |
| "learning_rate": 2.2931954636424284e-05, | |
| "loss": 1.9742, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.1508502468458585, | |
| "grad_norm": 0.7747183442115784, | |
| "learning_rate": 2.2723482321547697e-05, | |
| "loss": 1.9774, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.16456390565002743, | |
| "grad_norm": 0.8771799802780151, | |
| "learning_rate": 2.2515010006671114e-05, | |
| "loss": 1.9537, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17827756445419637, | |
| "grad_norm": 0.7963676452636719, | |
| "learning_rate": 2.230653769179453e-05, | |
| "loss": 1.9408, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.19199122325836535, | |
| "grad_norm": 0.7438392639160156, | |
| "learning_rate": 2.2098065376917944e-05, | |
| "loss": 2.0012, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2057048820625343, | |
| "grad_norm": 0.7908737063407898, | |
| "learning_rate": 2.188959306204136e-05, | |
| "loss": 1.9485, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.21941854086670323, | |
| "grad_norm": 0.7139158248901367, | |
| "learning_rate": 2.1681120747164777e-05, | |
| "loss": 1.9829, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23313219967087218, | |
| "grad_norm": 0.9640923738479614, | |
| "learning_rate": 2.147264843228819e-05, | |
| "loss": 2.0043, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.24684585847504115, | |
| "grad_norm": 0.8830370903015137, | |
| "learning_rate": 2.1264176117411607e-05, | |
| "loss": 1.9798, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.2605595172792101, | |
| "grad_norm": 0.8050561547279358, | |
| "learning_rate": 2.1055703802535024e-05, | |
| "loss": 1.9679, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.27427317608337903, | |
| "grad_norm": 0.8845964074134827, | |
| "learning_rate": 2.084723148765844e-05, | |
| "loss": 1.9261, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.287986834887548, | |
| "grad_norm": 0.7687380909919739, | |
| "learning_rate": 2.0638759172781854e-05, | |
| "loss": 1.938, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.301700493691717, | |
| "grad_norm": 0.8741185665130615, | |
| "learning_rate": 2.043028685790527e-05, | |
| "loss": 1.941, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3154141524958859, | |
| "grad_norm": 0.8414379358291626, | |
| "learning_rate": 2.0221814543028687e-05, | |
| "loss": 1.9442, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.32912781130005486, | |
| "grad_norm": 0.8874805569648743, | |
| "learning_rate": 2.00133422281521e-05, | |
| "loss": 1.9899, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.34284147010422383, | |
| "grad_norm": 0.8438735604286194, | |
| "learning_rate": 1.9804869913275517e-05, | |
| "loss": 1.9332, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.35655512890839275, | |
| "grad_norm": 0.8141827583312988, | |
| "learning_rate": 1.9596397598398934e-05, | |
| "loss": 1.9334, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3702687877125617, | |
| "grad_norm": 0.8998775482177734, | |
| "learning_rate": 1.9387925283522347e-05, | |
| "loss": 1.9208, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.3839824465167307, | |
| "grad_norm": 0.8827155232429504, | |
| "learning_rate": 1.9179452968645764e-05, | |
| "loss": 1.9287, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3976961053208996, | |
| "grad_norm": 0.7951996922492981, | |
| "learning_rate": 1.897098065376918e-05, | |
| "loss": 1.9514, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.4114097641250686, | |
| "grad_norm": 0.784051775932312, | |
| "learning_rate": 1.8762508338892594e-05, | |
| "loss": 1.9226, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4251234229292375, | |
| "grad_norm": 0.9527340531349182, | |
| "learning_rate": 1.855403602401601e-05, | |
| "loss": 1.9459, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.43883708173340646, | |
| "grad_norm": 0.8100705742835999, | |
| "learning_rate": 1.8345563709139427e-05, | |
| "loss": 1.9218, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.45255074053757544, | |
| "grad_norm": 0.814917266368866, | |
| "learning_rate": 1.8137091394262844e-05, | |
| "loss": 1.9682, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.46626439934174435, | |
| "grad_norm": 0.8693529367446899, | |
| "learning_rate": 1.7928619079386257e-05, | |
| "loss": 1.9068, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.4799780581459133, | |
| "grad_norm": 0.8571690320968628, | |
| "learning_rate": 1.7720146764509674e-05, | |
| "loss": 1.9225, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.4936917169500823, | |
| "grad_norm": 0.9239374995231628, | |
| "learning_rate": 1.751167444963309e-05, | |
| "loss": 1.9628, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5074053757542513, | |
| "grad_norm": 0.8887170553207397, | |
| "learning_rate": 1.7303202134756504e-05, | |
| "loss": 1.8972, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.5211190345584202, | |
| "grad_norm": 0.8989230394363403, | |
| "learning_rate": 1.709472981987992e-05, | |
| "loss": 1.9517, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5348326933625891, | |
| "grad_norm": 0.8530762791633606, | |
| "learning_rate": 1.6886257505003337e-05, | |
| "loss": 1.9163, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.5485463521667581, | |
| "grad_norm": 0.9289182424545288, | |
| "learning_rate": 1.667778519012675e-05, | |
| "loss": 1.9399, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5485463521667581, | |
| "eval_loss": 1.8345812559127808, | |
| "eval_runtime": 12.1751, | |
| "eval_samples_per_second": 11.91, | |
| "eval_steps_per_second": 1.561, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.562260010970927, | |
| "grad_norm": 0.9478083848953247, | |
| "learning_rate": 1.6469312875250167e-05, | |
| "loss": 1.9219, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.575973669775096, | |
| "grad_norm": 0.9554803371429443, | |
| "learning_rate": 1.6260840560373584e-05, | |
| "loss": 1.9297, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.589687328579265, | |
| "grad_norm": 0.9292556643486023, | |
| "learning_rate": 1.6052368245496997e-05, | |
| "loss": 1.9245, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.603400987383434, | |
| "grad_norm": 0.9070860743522644, | |
| "learning_rate": 1.5843895930620414e-05, | |
| "loss": 1.9247, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6171146461876028, | |
| "grad_norm": 0.883678674697876, | |
| "learning_rate": 1.563542361574383e-05, | |
| "loss": 1.9041, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.6308283049917718, | |
| "grad_norm": 0.9620354175567627, | |
| "learning_rate": 1.5426951300867247e-05, | |
| "loss": 1.9394, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.6445419637959408, | |
| "grad_norm": 0.9458388090133667, | |
| "learning_rate": 1.521847898599066e-05, | |
| "loss": 1.9513, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.6582556226001097, | |
| "grad_norm": 0.9318162798881531, | |
| "learning_rate": 1.5010006671114077e-05, | |
| "loss": 1.8923, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6719692814042787, | |
| "grad_norm": 0.8988519906997681, | |
| "learning_rate": 1.4801534356237492e-05, | |
| "loss": 1.9337, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.6856829402084477, | |
| "grad_norm": 0.9615154266357422, | |
| "learning_rate": 1.4593062041360909e-05, | |
| "loss": 1.9118, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6993965990126165, | |
| "grad_norm": 0.9634252190589905, | |
| "learning_rate": 1.4384589726484324e-05, | |
| "loss": 1.8784, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.7131102578167855, | |
| "grad_norm": 0.9005519151687622, | |
| "learning_rate": 1.4176117411607739e-05, | |
| "loss": 1.9124, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7268239166209545, | |
| "grad_norm": 0.9393877387046814, | |
| "learning_rate": 1.3967645096731155e-05, | |
| "loss": 1.925, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.7405375754251234, | |
| "grad_norm": 0.9061549305915833, | |
| "learning_rate": 1.375917278185457e-05, | |
| "loss": 1.9245, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7542512342292924, | |
| "grad_norm": 0.9310101270675659, | |
| "learning_rate": 1.3550700466977987e-05, | |
| "loss": 1.9177, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.7679648930334614, | |
| "grad_norm": 0.892022430896759, | |
| "learning_rate": 1.3342228152101402e-05, | |
| "loss": 1.8819, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7816785518376302, | |
| "grad_norm": 1.0094044208526611, | |
| "learning_rate": 1.3133755837224817e-05, | |
| "loss": 1.9032, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.7953922106417992, | |
| "grad_norm": 0.9686225652694702, | |
| "learning_rate": 1.2925283522348234e-05, | |
| "loss": 1.9229, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.8091058694459682, | |
| "grad_norm": 0.8588367104530334, | |
| "learning_rate": 1.2716811207471649e-05, | |
| "loss": 1.8997, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.8228195282501372, | |
| "grad_norm": 0.9800122380256653, | |
| "learning_rate": 1.2508338892595064e-05, | |
| "loss": 1.9087, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8365331870543061, | |
| "grad_norm": 0.9647035002708435, | |
| "learning_rate": 1.229986657771848e-05, | |
| "loss": 1.9239, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.850246845858475, | |
| "grad_norm": 0.9459319114685059, | |
| "learning_rate": 1.2091394262841895e-05, | |
| "loss": 1.88, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.863960504662644, | |
| "grad_norm": 0.989025354385376, | |
| "learning_rate": 1.1882921947965312e-05, | |
| "loss": 1.9134, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.8776741634668129, | |
| "grad_norm": 0.9573891162872314, | |
| "learning_rate": 1.1674449633088725e-05, | |
| "loss": 1.917, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8913878222709819, | |
| "grad_norm": 0.9619189500808716, | |
| "learning_rate": 1.1465977318212142e-05, | |
| "loss": 1.9202, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.9051014810751509, | |
| "grad_norm": 0.959064781665802, | |
| "learning_rate": 1.1257505003335557e-05, | |
| "loss": 1.9053, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.9188151398793198, | |
| "grad_norm": 1.0735059976577759, | |
| "learning_rate": 1.1049032688458972e-05, | |
| "loss": 1.914, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.9325287986834887, | |
| "grad_norm": 0.8986095786094666, | |
| "learning_rate": 1.0840560373582389e-05, | |
| "loss": 1.8883, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9462424574876577, | |
| "grad_norm": 0.9695867300033569, | |
| "learning_rate": 1.0632088058705804e-05, | |
| "loss": 1.892, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.9599561162918266, | |
| "grad_norm": 0.893974244594574, | |
| "learning_rate": 1.042361574382922e-05, | |
| "loss": 1.9094, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9736697750959956, | |
| "grad_norm": 1.0204190015792847, | |
| "learning_rate": 1.0215143428952635e-05, | |
| "loss": 1.9121, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.9873834339001646, | |
| "grad_norm": 0.9737017750740051, | |
| "learning_rate": 1.000667111407605e-05, | |
| "loss": 1.9003, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.0010970927043334, | |
| "grad_norm": 0.9492760300636292, | |
| "learning_rate": 9.798198799199467e-06, | |
| "loss": 1.9097, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.0148107515085025, | |
| "grad_norm": 0.9199559688568115, | |
| "learning_rate": 9.589726484322882e-06, | |
| "loss": 1.7813, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.0285244103126714, | |
| "grad_norm": 0.8272762298583984, | |
| "learning_rate": 9.381254169446297e-06, | |
| "loss": 1.8073, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.0422380691168405, | |
| "grad_norm": 0.9935997724533081, | |
| "learning_rate": 9.172781854569714e-06, | |
| "loss": 1.7889, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.0559517279210093, | |
| "grad_norm": 1.0486711263656616, | |
| "learning_rate": 8.964309539693129e-06, | |
| "loss": 1.8072, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.0696653867251782, | |
| "grad_norm": 1.0662914514541626, | |
| "learning_rate": 8.755837224816545e-06, | |
| "loss": 1.8171, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.0833790455293473, | |
| "grad_norm": 0.9916715621948242, | |
| "learning_rate": 8.54736490993996e-06, | |
| "loss": 1.8002, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.0970927043335161, | |
| "grad_norm": 1.0449947118759155, | |
| "learning_rate": 8.338892595063375e-06, | |
| "loss": 1.8416, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0970927043335161, | |
| "eval_loss": 1.7659664154052734, | |
| "eval_runtime": 12.173, | |
| "eval_samples_per_second": 11.912, | |
| "eval_steps_per_second": 1.561, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.1108063631376852, | |
| "grad_norm": 1.0766034126281738, | |
| "learning_rate": 8.130420280186792e-06, | |
| "loss": 1.8049, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.124520021941854, | |
| "grad_norm": 1.1313064098358154, | |
| "learning_rate": 7.921947965310207e-06, | |
| "loss": 1.7966, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.1382336807460232, | |
| "grad_norm": 1.219266653060913, | |
| "learning_rate": 7.713475650433624e-06, | |
| "loss": 1.839, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.151947339550192, | |
| "grad_norm": 1.171193242073059, | |
| "learning_rate": 7.5050033355570386e-06, | |
| "loss": 1.8187, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.1656609983543609, | |
| "grad_norm": 1.0916893482208252, | |
| "learning_rate": 7.296531020680454e-06, | |
| "loss": 1.8144, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 1.17937465715853, | |
| "grad_norm": 1.0833641290664673, | |
| "learning_rate": 7.088058705803869e-06, | |
| "loss": 1.8367, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.1930883159626988, | |
| "grad_norm": 1.1324750185012817, | |
| "learning_rate": 6.879586390927285e-06, | |
| "loss": 1.8323, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 1.2068019747668677, | |
| "grad_norm": 1.1888611316680908, | |
| "learning_rate": 6.671114076050701e-06, | |
| "loss": 1.8014, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.2205156335710368, | |
| "grad_norm": 1.2182481288909912, | |
| "learning_rate": 6.462641761174117e-06, | |
| "loss": 1.8353, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 1.2342292923752056, | |
| "grad_norm": 1.2250556945800781, | |
| "learning_rate": 6.254169446297532e-06, | |
| "loss": 1.816, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.2479429511793747, | |
| "grad_norm": 1.2347311973571777, | |
| "learning_rate": 6.045697131420948e-06, | |
| "loss": 1.8175, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 1.2616566099835436, | |
| "grad_norm": 1.2559269666671753, | |
| "learning_rate": 5.837224816544363e-06, | |
| "loss": 1.8196, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.2753702687877126, | |
| "grad_norm": 1.2170192003250122, | |
| "learning_rate": 5.6287525016677785e-06, | |
| "loss": 1.8137, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 1.2890839275918815, | |
| "grad_norm": 1.2384990453720093, | |
| "learning_rate": 5.420280186791194e-06, | |
| "loss": 1.7757, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.3027975863960504, | |
| "grad_norm": 1.2390954494476318, | |
| "learning_rate": 5.21180787191461e-06, | |
| "loss": 1.7952, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 1.3165112452002194, | |
| "grad_norm": 1.3812655210494995, | |
| "learning_rate": 5.003335557038025e-06, | |
| "loss": 1.8311, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.3302249040043883, | |
| "grad_norm": 1.2410950660705566, | |
| "learning_rate": 4.794863242161441e-06, | |
| "loss": 1.7996, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 1.3439385628085574, | |
| "grad_norm": 1.1785918474197388, | |
| "learning_rate": 4.586390927284857e-06, | |
| "loss": 1.793, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.3576522216127263, | |
| "grad_norm": 1.4189759492874146, | |
| "learning_rate": 4.377918612408273e-06, | |
| "loss": 1.8558, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 1.3713658804168953, | |
| "grad_norm": 1.209304928779602, | |
| "learning_rate": 4.169446297531688e-06, | |
| "loss": 1.7649, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.3850795392210642, | |
| "grad_norm": 1.2057141065597534, | |
| "learning_rate": 3.9609739826551035e-06, | |
| "loss": 1.8148, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 1.398793198025233, | |
| "grad_norm": 1.0700486898422241, | |
| "learning_rate": 3.7525016677785193e-06, | |
| "loss": 1.8063, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.4125068568294021, | |
| "grad_norm": 1.1492820978164673, | |
| "learning_rate": 3.5440293529019347e-06, | |
| "loss": 1.7777, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 1.426220515633571, | |
| "grad_norm": 1.2483413219451904, | |
| "learning_rate": 3.3355570380253505e-06, | |
| "loss": 1.8595, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.4399341744377399, | |
| "grad_norm": 1.384775996208191, | |
| "learning_rate": 3.127084723148766e-06, | |
| "loss": 1.7923, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 1.453647833241909, | |
| "grad_norm": 1.2097506523132324, | |
| "learning_rate": 2.9186124082721813e-06, | |
| "loss": 1.8107, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.467361492046078, | |
| "grad_norm": 1.2465295791625977, | |
| "learning_rate": 2.710140093395597e-06, | |
| "loss": 1.8037, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 1.4810751508502469, | |
| "grad_norm": 1.3312894105911255, | |
| "learning_rate": 2.5016677785190126e-06, | |
| "loss": 1.8149, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.4947888096544157, | |
| "grad_norm": 1.1086313724517822, | |
| "learning_rate": 2.2931954636424284e-06, | |
| "loss": 1.8261, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 1.5085024684585848, | |
| "grad_norm": 1.3292433023452759, | |
| "learning_rate": 2.084723148765844e-06, | |
| "loss": 1.8145, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.5222161272627537, | |
| "grad_norm": 1.2049185037612915, | |
| "learning_rate": 1.8762508338892596e-06, | |
| "loss": 1.8255, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 1.5359297860669225, | |
| "grad_norm": 1.266183614730835, | |
| "learning_rate": 1.6677785190126753e-06, | |
| "loss": 1.7954, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.5496434448710916, | |
| "grad_norm": 1.337235927581787, | |
| "learning_rate": 1.4593062041360907e-06, | |
| "loss": 1.7749, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 1.5633571036752607, | |
| "grad_norm": 1.3112897872924805, | |
| "learning_rate": 1.2508338892595063e-06, | |
| "loss": 1.8236, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.5770707624794296, | |
| "grad_norm": 1.24995756149292, | |
| "learning_rate": 1.042361574382922e-06, | |
| "loss": 1.8364, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 1.5907844212835984, | |
| "grad_norm": 1.2855443954467773, | |
| "learning_rate": 8.338892595063376e-07, | |
| "loss": 1.8267, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.6044980800877675, | |
| "grad_norm": 1.1317636966705322, | |
| "learning_rate": 6.254169446297531e-07, | |
| "loss": 1.8185, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 1.6182117388919364, | |
| "grad_norm": 1.3289167881011963, | |
| "learning_rate": 4.169446297531688e-07, | |
| "loss": 1.8577, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.6319253976961052, | |
| "grad_norm": 1.2797874212265015, | |
| "learning_rate": 2.084723148765844e-07, | |
| "loss": 1.811, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 1.6456390565002743, | |
| "grad_norm": 1.3882725238800049, | |
| "learning_rate": 0.0, | |
| "loss": 1.8245, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.6456390565002743, | |
| "eval_loss": 1.7348041534423828, | |
| "eval_runtime": 12.1692, | |
| "eval_samples_per_second": 11.915, | |
| "eval_steps_per_second": 1.561, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 3000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0737917404957901e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |