| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.2, |
| "eval_steps": 500, |
| "global_step": 400, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.008, |
| "grad_norm": 0.09337692707777023, |
| "learning_rate": 2e-05, |
| "loss": 0.015294969081878662, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.016, |
| "grad_norm": 0.07111852616071701, |
| "learning_rate": 2e-05, |
| "loss": 0.007340883836150169, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.024, |
| "grad_norm": 0.07195291668176651, |
| "learning_rate": 2e-05, |
| "loss": 0.0077208466827869415, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.032, |
| "grad_norm": 0.09681837260723114, |
| "learning_rate": 2e-05, |
| "loss": 0.008473804220557213, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.04, |
| "grad_norm": 0.06043309345841408, |
| "learning_rate": 2e-05, |
| "loss": 0.006642022170126438, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.048, |
| "grad_norm": 0.05604363605380058, |
| "learning_rate": 2e-05, |
| "loss": 0.003878322197124362, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.056, |
| "grad_norm": 0.06847543269395828, |
| "learning_rate": 2e-05, |
| "loss": 0.011454813182353973, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.064, |
| "grad_norm": 0.05413937568664551, |
| "learning_rate": 2e-05, |
| "loss": 0.005098872818052769, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.072, |
| "grad_norm": 0.04113985598087311, |
| "learning_rate": 2e-05, |
| "loss": 0.0072427294217050076, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.08, |
| "grad_norm": 0.06007072329521179, |
| "learning_rate": 2e-05, |
| "loss": 0.005184551235288382, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.088, |
| "grad_norm": 0.05269957333803177, |
| "learning_rate": 2e-05, |
| "loss": 0.006402506493031979, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.096, |
| "grad_norm": 0.07171762734651566, |
| "learning_rate": 2e-05, |
| "loss": 0.006807660683989525, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.104, |
| "grad_norm": 0.04405634105205536, |
| "learning_rate": 2e-05, |
| "loss": 0.0049944375641644, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.112, |
| "grad_norm": 0.04304119199514389, |
| "learning_rate": 2e-05, |
| "loss": 0.005410908255726099, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.12, |
| "grad_norm": 0.057690370827913284, |
| "learning_rate": 2e-05, |
| "loss": 0.007783032953739166, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.128, |
| "grad_norm": 0.04541298747062683, |
| "learning_rate": 2e-05, |
| "loss": 0.004243949428200722, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.136, |
| "grad_norm": 0.06319569051265717, |
| "learning_rate": 2e-05, |
| "loss": 0.005336169619113207, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.144, |
| "grad_norm": 0.03583596274256706, |
| "learning_rate": 2e-05, |
| "loss": 0.004895349964499474, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.152, |
| "grad_norm": 0.05986885726451874, |
| "learning_rate": 2e-05, |
| "loss": 0.006950004491955042, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.16, |
| "grad_norm": 0.08991007506847382, |
| "learning_rate": 2e-05, |
| "loss": 0.00810159184038639, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.168, |
| "grad_norm": 0.05890341475605965, |
| "learning_rate": 2e-05, |
| "loss": 0.006250896491110325, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.176, |
| "grad_norm": 0.04574795812368393, |
| "learning_rate": 2e-05, |
| "loss": 0.005519284401088953, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.184, |
| "grad_norm": 0.05631595849990845, |
| "learning_rate": 2e-05, |
| "loss": 0.00560405757278204, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.192, |
| "grad_norm": 0.06935696303844452, |
| "learning_rate": 2e-05, |
| "loss": 0.00864097848534584, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.2, |
| "grad_norm": 0.04360036551952362, |
| "learning_rate": 2e-05, |
| "loss": 0.003752000629901886, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.208, |
| "grad_norm": 0.05469367280602455, |
| "learning_rate": 2e-05, |
| "loss": 0.0067576696164906025, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.216, |
| "grad_norm": 0.06384038180112839, |
| "learning_rate": 2e-05, |
| "loss": 0.004895553924143314, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.224, |
| "grad_norm": 0.06384114921092987, |
| "learning_rate": 2e-05, |
| "loss": 0.005900057964026928, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.232, |
| "grad_norm": 0.0655912533402443, |
| "learning_rate": 2e-05, |
| "loss": 0.005855287425220013, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.24, |
| "grad_norm": 0.04619958996772766, |
| "learning_rate": 2e-05, |
| "loss": 0.005466114729642868, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.248, |
| "grad_norm": 0.04525391384959221, |
| "learning_rate": 2e-05, |
| "loss": 0.004368640948086977, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.256, |
| "grad_norm": 0.04429985210299492, |
| "learning_rate": 2e-05, |
| "loss": 0.005769067443907261, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.264, |
| "grad_norm": 0.03835798799991608, |
| "learning_rate": 2e-05, |
| "loss": 0.004946814849972725, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.272, |
| "grad_norm": 0.06351771950721741, |
| "learning_rate": 2e-05, |
| "loss": 0.005820239428430796, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.28, |
| "grad_norm": 0.05175631865859032, |
| "learning_rate": 2e-05, |
| "loss": 0.007221951149404049, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.288, |
| "grad_norm": 0.031039636582136154, |
| "learning_rate": 2e-05, |
| "loss": 0.0024754288606345654, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.296, |
| "grad_norm": 0.04708324000239372, |
| "learning_rate": 2e-05, |
| "loss": 0.002840370638296008, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.304, |
| "grad_norm": 0.06475556641817093, |
| "learning_rate": 2e-05, |
| "loss": 0.0077974689193069935, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.312, |
| "grad_norm": 0.048895902931690216, |
| "learning_rate": 2e-05, |
| "loss": 0.0040883091278374195, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.32, |
| "grad_norm": 0.058496665209531784, |
| "learning_rate": 2e-05, |
| "loss": 0.004431126173585653, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.328, |
| "grad_norm": 0.0752127543091774, |
| "learning_rate": 2e-05, |
| "loss": 0.009588265791535378, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.336, |
| "grad_norm": 0.029396817088127136, |
| "learning_rate": 2e-05, |
| "loss": 0.004547424148768187, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.344, |
| "grad_norm": 0.10113711655139923, |
| "learning_rate": 2e-05, |
| "loss": 0.00836176797747612, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.352, |
| "grad_norm": 0.04212963953614235, |
| "learning_rate": 2e-05, |
| "loss": 0.004907882306724787, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.36, |
| "grad_norm": 0.02902965061366558, |
| "learning_rate": 2e-05, |
| "loss": 0.002945390995591879, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.368, |
| "grad_norm": 0.05033962428569794, |
| "learning_rate": 2e-05, |
| "loss": 0.005018638446927071, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.376, |
| "grad_norm": 0.034906402230262756, |
| "learning_rate": 2e-05, |
| "loss": 0.0026375213637948036, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.384, |
| "grad_norm": 0.03871266171336174, |
| "learning_rate": 2e-05, |
| "loss": 0.004687938839197159, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.392, |
| "grad_norm": 0.040328506380319595, |
| "learning_rate": 2e-05, |
| "loss": 0.006280779838562012, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 0.04712209478020668, |
| "learning_rate": 2e-05, |
| "loss": 0.0031323714647442102, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.408, |
| "grad_norm": 0.045797791332006454, |
| "learning_rate": 2e-05, |
| "loss": 0.004443546291440725, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.416, |
| "grad_norm": 0.06185959652066231, |
| "learning_rate": 2e-05, |
| "loss": 0.008573911152780056, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.424, |
| "grad_norm": 0.059398140758275986, |
| "learning_rate": 2e-05, |
| "loss": 0.00942009873688221, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.432, |
| "grad_norm": 0.07633177191019058, |
| "learning_rate": 2e-05, |
| "loss": 0.009616912342607975, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.44, |
| "grad_norm": 0.053814005106687546, |
| "learning_rate": 2e-05, |
| "loss": 0.007603101897984743, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.448, |
| "grad_norm": 0.05707194656133652, |
| "learning_rate": 2e-05, |
| "loss": 0.005355853587388992, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.456, |
| "grad_norm": 0.04392722621560097, |
| "learning_rate": 2e-05, |
| "loss": 0.003092748112976551, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.464, |
| "grad_norm": 0.027375129982829094, |
| "learning_rate": 2e-05, |
| "loss": 0.003245085943490267, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.472, |
| "grad_norm": 0.05855637416243553, |
| "learning_rate": 2e-05, |
| "loss": 0.006621942389756441, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.48, |
| "grad_norm": 0.040302883833646774, |
| "learning_rate": 2e-05, |
| "loss": 0.004884965717792511, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.488, |
| "grad_norm": 0.038790635764598846, |
| "learning_rate": 2e-05, |
| "loss": 0.0041805170476436615, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.496, |
| "grad_norm": 0.059503670781850815, |
| "learning_rate": 2e-05, |
| "loss": 0.0071491943672299385, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.504, |
| "grad_norm": 0.059957440942525864, |
| "learning_rate": 2e-05, |
| "loss": 0.006595572456717491, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.512, |
| "grad_norm": 0.05570148676633835, |
| "learning_rate": 2e-05, |
| "loss": 0.007463263813406229, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.52, |
| "grad_norm": 0.056187067180871964, |
| "learning_rate": 2e-05, |
| "loss": 0.004543005023151636, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.528, |
| "grad_norm": 0.04404418170452118, |
| "learning_rate": 2e-05, |
| "loss": 0.006366059184074402, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.536, |
| "grad_norm": 0.049447719007730484, |
| "learning_rate": 2e-05, |
| "loss": 0.006581606809049845, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.544, |
| "grad_norm": 0.03419354185461998, |
| "learning_rate": 2e-05, |
| "loss": 0.0035948362201452255, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.552, |
| "grad_norm": 0.05780700221657753, |
| "learning_rate": 2e-05, |
| "loss": 0.005125825759023428, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.56, |
| "grad_norm": 0.06544717401266098, |
| "learning_rate": 2e-05, |
| "loss": 0.006060135085135698, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.568, |
| "grad_norm": 0.04413054138422012, |
| "learning_rate": 2e-05, |
| "loss": 0.003213978838175535, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.576, |
| "grad_norm": 0.039240192621946335, |
| "learning_rate": 2e-05, |
| "loss": 0.004292209167033434, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.584, |
| "grad_norm": 0.060329239815473557, |
| "learning_rate": 2e-05, |
| "loss": 0.006161934696137905, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.592, |
| "grad_norm": 0.06489120423793793, |
| "learning_rate": 2e-05, |
| "loss": 0.0032440992072224617, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.6, |
| "grad_norm": 0.04400647059082985, |
| "learning_rate": 2e-05, |
| "loss": 0.004546411335468292, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.608, |
| "grad_norm": 0.046490587294101715, |
| "learning_rate": 2e-05, |
| "loss": 0.00506614800542593, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.616, |
| "grad_norm": 0.03474031388759613, |
| "learning_rate": 2e-05, |
| "loss": 0.003143966430798173, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.624, |
| "grad_norm": 0.04975699633359909, |
| "learning_rate": 2e-05, |
| "loss": 0.004066984634846449, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.632, |
| "grad_norm": 0.06441677361726761, |
| "learning_rate": 2e-05, |
| "loss": 0.007056248374283314, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.64, |
| "grad_norm": 0.05794770270586014, |
| "learning_rate": 2e-05, |
| "loss": 0.0037886211648583412, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.648, |
| "grad_norm": 0.06725575029850006, |
| "learning_rate": 2e-05, |
| "loss": 0.005529267713427544, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.656, |
| "grad_norm": 0.05235245078802109, |
| "learning_rate": 2e-05, |
| "loss": 0.0031472702976316214, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.664, |
| "grad_norm": 0.0473208874464035, |
| "learning_rate": 2e-05, |
| "loss": 0.004731162916868925, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.672, |
| "grad_norm": 0.05135403573513031, |
| "learning_rate": 2e-05, |
| "loss": 0.005146307405084372, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.68, |
| "grad_norm": 0.04229608178138733, |
| "learning_rate": 2e-05, |
| "loss": 0.004758044611662626, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.688, |
| "grad_norm": 0.04035644605755806, |
| "learning_rate": 2e-05, |
| "loss": 0.0043296488001942635, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.696, |
| "grad_norm": 0.060147274285554886, |
| "learning_rate": 2e-05, |
| "loss": 0.008551320061087608, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.704, |
| "grad_norm": 0.0395248718559742, |
| "learning_rate": 2e-05, |
| "loss": 0.0024903551675379276, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.712, |
| "grad_norm": 0.071822389960289, |
| "learning_rate": 2e-05, |
| "loss": 0.008494703099131584, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.72, |
| "grad_norm": 0.058960266411304474, |
| "learning_rate": 2e-05, |
| "loss": 0.006603310350328684, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.728, |
| "grad_norm": 0.05052534490823746, |
| "learning_rate": 2e-05, |
| "loss": 0.0058289458975195885, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.736, |
| "grad_norm": 0.049510154873132706, |
| "learning_rate": 2e-05, |
| "loss": 0.003644926007837057, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.744, |
| "grad_norm": 0.0739966556429863, |
| "learning_rate": 2e-05, |
| "loss": 0.0028007840737700462, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.752, |
| "grad_norm": 0.07646612823009491, |
| "learning_rate": 2e-05, |
| "loss": 0.010534452274441719, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.76, |
| "grad_norm": 0.05939766764640808, |
| "learning_rate": 2e-05, |
| "loss": 0.004194422625005245, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.768, |
| "grad_norm": 0.049664683640003204, |
| "learning_rate": 2e-05, |
| "loss": 0.003597400849685073, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.776, |
| "grad_norm": 0.13988253474235535, |
| "learning_rate": 2e-05, |
| "loss": 0.006600011605769396, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.784, |
| "grad_norm": 0.06141381338238716, |
| "learning_rate": 2e-05, |
| "loss": 0.004176464397460222, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.792, |
| "grad_norm": 0.040125612169504166, |
| "learning_rate": 2e-05, |
| "loss": 0.0029954789206385612, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 0.05379870533943176, |
| "learning_rate": 2e-05, |
| "loss": 0.005120614543557167, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.808, |
| "grad_norm": 0.04878966137766838, |
| "learning_rate": 2e-05, |
| "loss": 0.0049518863670527935, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.816, |
| "grad_norm": 0.0832071602344513, |
| "learning_rate": 2e-05, |
| "loss": 0.003636852605268359, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.824, |
| "grad_norm": 0.1109263077378273, |
| "learning_rate": 2e-05, |
| "loss": 0.011921115219593048, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.832, |
| "grad_norm": 0.037266287952661514, |
| "learning_rate": 2e-05, |
| "loss": 0.002762642689049244, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.84, |
| "grad_norm": 0.044354166835546494, |
| "learning_rate": 2e-05, |
| "loss": 0.004988430067896843, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.848, |
| "grad_norm": 0.03570897877216339, |
| "learning_rate": 2e-05, |
| "loss": 0.002259196015074849, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.856, |
| "grad_norm": 0.059259530156850815, |
| "learning_rate": 2e-05, |
| "loss": 0.006721946410834789, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.864, |
| "grad_norm": 0.05184999480843544, |
| "learning_rate": 2e-05, |
| "loss": 0.005531419534236193, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.872, |
| "grad_norm": 0.08037418127059937, |
| "learning_rate": 2e-05, |
| "loss": 0.006799847353249788, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.88, |
| "grad_norm": 0.039056453853845596, |
| "learning_rate": 2e-05, |
| "loss": 0.00481954962015152, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.888, |
| "grad_norm": 0.05626671388745308, |
| "learning_rate": 2e-05, |
| "loss": 0.0032779546454548836, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.896, |
| "grad_norm": 0.05177405849099159, |
| "learning_rate": 2e-05, |
| "loss": 0.004294276237487793, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.904, |
| "grad_norm": 0.029522892087697983, |
| "learning_rate": 2e-05, |
| "loss": 0.0030598388984799385, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.912, |
| "grad_norm": 0.05130866542458534, |
| "learning_rate": 2e-05, |
| "loss": 0.005800556857138872, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.92, |
| "grad_norm": 0.021749891340732574, |
| "learning_rate": 2e-05, |
| "loss": 0.001492157345637679, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.928, |
| "grad_norm": 0.0728369727730751, |
| "learning_rate": 2e-05, |
| "loss": 0.008168019354343414, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.936, |
| "grad_norm": 0.051655277609825134, |
| "learning_rate": 2e-05, |
| "loss": 0.004517991095781326, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.944, |
| "grad_norm": 0.028527911752462387, |
| "learning_rate": 2e-05, |
| "loss": 0.004356182646006346, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.952, |
| "grad_norm": 0.04092099890112877, |
| "learning_rate": 2e-05, |
| "loss": 0.0035778400488197803, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.96, |
| "grad_norm": 0.06751365959644318, |
| "learning_rate": 2e-05, |
| "loss": 0.006765933241695166, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.968, |
| "grad_norm": 0.06141749769449234, |
| "learning_rate": 2e-05, |
| "loss": 0.010381967760622501, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.976, |
| "grad_norm": 0.0584125854074955, |
| "learning_rate": 2e-05, |
| "loss": 0.0059956214390695095, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.984, |
| "grad_norm": 0.06699343770742416, |
| "learning_rate": 2e-05, |
| "loss": 0.006007581949234009, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.992, |
| "grad_norm": 0.04457642510533333, |
| "learning_rate": 2e-05, |
| "loss": 0.00433978158980608, |
| "step": 124 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.0724361315369606, |
| "learning_rate": 2e-05, |
| "loss": 0.009294194169342518, |
| "step": 125 |
| }, |
| { |
| "epoch": 1.008, |
| "grad_norm": 0.041911620646715164, |
| "learning_rate": 2e-05, |
| "loss": 0.0031240973621606827, |
| "step": 126 |
| }, |
| { |
| "epoch": 1.016, |
| "grad_norm": 0.03944513201713562, |
| "learning_rate": 2e-05, |
| "loss": 0.005666844546794891, |
| "step": 127 |
| }, |
| { |
| "epoch": 1.024, |
| "grad_norm": 0.024236207827925682, |
| "learning_rate": 2e-05, |
| "loss": 0.002599143423140049, |
| "step": 128 |
| }, |
| { |
| "epoch": 1.032, |
| "grad_norm": 0.02648847922682762, |
| "learning_rate": 2e-05, |
| "loss": 0.002489139325916767, |
| "step": 129 |
| }, |
| { |
| "epoch": 1.04, |
| "grad_norm": 0.03522297367453575, |
| "learning_rate": 2e-05, |
| "loss": 0.0031560207717120647, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.048, |
| "grad_norm": 0.03963733837008476, |
| "learning_rate": 2e-05, |
| "loss": 0.002944512525573373, |
| "step": 131 |
| }, |
| { |
| "epoch": 1.056, |
| "grad_norm": 0.043245717883110046, |
| "learning_rate": 2e-05, |
| "loss": 0.005929629784077406, |
| "step": 132 |
| }, |
| { |
| "epoch": 1.064, |
| "grad_norm": 0.04024028778076172, |
| "learning_rate": 2e-05, |
| "loss": 0.0028088355902582407, |
| "step": 133 |
| }, |
| { |
| "epoch": 1.072, |
| "grad_norm": 0.03739921376109123, |
| "learning_rate": 2e-05, |
| "loss": 0.0025795348919928074, |
| "step": 134 |
| }, |
| { |
| "epoch": 1.08, |
| "grad_norm": 0.053612496703863144, |
| "learning_rate": 2e-05, |
| "loss": 0.0051528215408325195, |
| "step": 135 |
| }, |
| { |
| "epoch": 1.088, |
| "grad_norm": 0.03874121978878975, |
| "learning_rate": 2e-05, |
| "loss": 0.0029356672894209623, |
| "step": 136 |
| }, |
| { |
| "epoch": 1.096, |
| "grad_norm": 0.04209478199481964, |
| "learning_rate": 2e-05, |
| "loss": 0.004753364250063896, |
| "step": 137 |
| }, |
| { |
| "epoch": 1.104, |
| "grad_norm": 0.032321907579898834, |
| "learning_rate": 2e-05, |
| "loss": 0.002570532029494643, |
| "step": 138 |
| }, |
| { |
| "epoch": 1.112, |
| "grad_norm": 0.03770575299859047, |
| "learning_rate": 2e-05, |
| "loss": 0.0032067985739558935, |
| "step": 139 |
| }, |
| { |
| "epoch": 1.12, |
| "grad_norm": 0.042145032435655594, |
| "learning_rate": 2e-05, |
| "loss": 0.004715348593890667, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.1280000000000001, |
| "grad_norm": 0.04259483888745308, |
| "learning_rate": 2e-05, |
| "loss": 0.0040532853454351425, |
| "step": 141 |
| }, |
| { |
| "epoch": 1.1360000000000001, |
| "grad_norm": 0.04432355985045433, |
| "learning_rate": 2e-05, |
| "loss": 0.004878481850028038, |
| "step": 142 |
| }, |
| { |
| "epoch": 1.144, |
| "grad_norm": 0.04144922271370888, |
| "learning_rate": 2e-05, |
| "loss": 0.0035966027062386274, |
| "step": 143 |
| }, |
| { |
| "epoch": 1.152, |
| "grad_norm": 0.034188415855169296, |
| "learning_rate": 2e-05, |
| "loss": 0.0035179737024009228, |
| "step": 144 |
| }, |
| { |
| "epoch": 1.16, |
| "grad_norm": 0.04314472898840904, |
| "learning_rate": 2e-05, |
| "loss": 0.00275869807228446, |
| "step": 145 |
| }, |
| { |
| "epoch": 1.168, |
| "grad_norm": 0.03345927596092224, |
| "learning_rate": 2e-05, |
| "loss": 0.004454055335372686, |
| "step": 146 |
| }, |
| { |
| "epoch": 1.176, |
| "grad_norm": 0.032709989696741104, |
| "learning_rate": 2e-05, |
| "loss": 0.002561516361311078, |
| "step": 147 |
| }, |
| { |
| "epoch": 1.184, |
| "grad_norm": 0.037629563361406326, |
| "learning_rate": 2e-05, |
| "loss": 0.002726494800299406, |
| "step": 148 |
| }, |
| { |
| "epoch": 1.192, |
| "grad_norm": 0.03565853834152222, |
| "learning_rate": 2e-05, |
| "loss": 0.005714224185794592, |
| "step": 149 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 0.04101553559303284, |
| "learning_rate": 2e-05, |
| "loss": 0.00477135693654418, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.208, |
| "grad_norm": 0.02103736251592636, |
| "learning_rate": 2e-05, |
| "loss": 0.00214054505340755, |
| "step": 151 |
| }, |
| { |
| "epoch": 1.216, |
| "grad_norm": 0.038353558629751205, |
| "learning_rate": 2e-05, |
| "loss": 0.0023091635666787624, |
| "step": 152 |
| }, |
| { |
| "epoch": 1.224, |
| "grad_norm": 0.03039981611073017, |
| "learning_rate": 2e-05, |
| "loss": 0.002429583575576544, |
| "step": 153 |
| }, |
| { |
| "epoch": 1.232, |
| "grad_norm": 0.038060858845710754, |
| "learning_rate": 2e-05, |
| "loss": 0.0024876513052731752, |
| "step": 154 |
| }, |
| { |
| "epoch": 1.24, |
| "grad_norm": 0.10335491597652435, |
| "learning_rate": 2e-05, |
| "loss": 0.0025280544068664312, |
| "step": 155 |
| }, |
| { |
| "epoch": 1.248, |
| "grad_norm": 0.040877990424633026, |
| "learning_rate": 2e-05, |
| "loss": 0.004313280805945396, |
| "step": 156 |
| }, |
| { |
| "epoch": 1.256, |
| "grad_norm": 0.05562249571084976, |
| "learning_rate": 2e-05, |
| "loss": 0.0048119621351361275, |
| "step": 157 |
| }, |
| { |
| "epoch": 1.264, |
| "grad_norm": 0.05108984187245369, |
| "learning_rate": 2e-05, |
| "loss": 0.00518863694742322, |
| "step": 158 |
| }, |
| { |
| "epoch": 1.272, |
| "grad_norm": 0.05357106775045395, |
| "learning_rate": 2e-05, |
| "loss": 0.0024496042169630527, |
| "step": 159 |
| }, |
| { |
| "epoch": 1.28, |
| "grad_norm": 0.037901945412158966, |
| "learning_rate": 2e-05, |
| "loss": 0.002475207205861807, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.288, |
| "grad_norm": 0.0420001782476902, |
| "learning_rate": 2e-05, |
| "loss": 0.0058222366496920586, |
| "step": 161 |
| }, |
| { |
| "epoch": 1.296, |
| "grad_norm": 0.05443611368536949, |
| "learning_rate": 2e-05, |
| "loss": 0.0052161430940032005, |
| "step": 162 |
| }, |
| { |
| "epoch": 1.304, |
| "grad_norm": 0.05211922898888588, |
| "learning_rate": 2e-05, |
| "loss": 0.004603986628353596, |
| "step": 163 |
| }, |
| { |
| "epoch": 1.312, |
| "grad_norm": 0.04726783558726311, |
| "learning_rate": 2e-05, |
| "loss": 0.004252967424690723, |
| "step": 164 |
| }, |
| { |
| "epoch": 1.32, |
| "grad_norm": 0.07070634514093399, |
| "learning_rate": 2e-05, |
| "loss": 0.006596374791115522, |
| "step": 165 |
| }, |
| { |
| "epoch": 1.328, |
| "grad_norm": 0.05988767370581627, |
| "learning_rate": 2e-05, |
| "loss": 0.006062244065105915, |
| "step": 166 |
| }, |
| { |
| "epoch": 1.336, |
| "grad_norm": 0.06943166255950928, |
| "learning_rate": 2e-05, |
| "loss": 0.007745644543319941, |
| "step": 167 |
| }, |
| { |
| "epoch": 1.3439999999999999, |
| "grad_norm": 0.03787314146757126, |
| "learning_rate": 2e-05, |
| "loss": 0.003922807518392801, |
| "step": 168 |
| }, |
| { |
| "epoch": 1.3519999999999999, |
| "grad_norm": 0.02878367155790329, |
| "learning_rate": 2e-05, |
| "loss": 0.0019241905538365245, |
| "step": 169 |
| }, |
| { |
| "epoch": 1.3599999999999999, |
| "grad_norm": 0.05218566954135895, |
| "learning_rate": 2e-05, |
| "loss": 0.0058232140727341175, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.3679999999999999, |
| "grad_norm": 0.06744556874036789, |
| "learning_rate": 2e-05, |
| "loss": 0.004936105106025934, |
| "step": 171 |
| }, |
| { |
| "epoch": 1.376, |
| "grad_norm": 0.04349220544099808, |
| "learning_rate": 2e-05, |
| "loss": 0.003136698855087161, |
| "step": 172 |
| }, |
| { |
| "epoch": 1.384, |
| "grad_norm": 0.046757787466049194, |
| "learning_rate": 2e-05, |
| "loss": 0.004662223160266876, |
| "step": 173 |
| }, |
| { |
| "epoch": 1.392, |
| "grad_norm": 0.03483404964208603, |
| "learning_rate": 2e-05, |
| "loss": 0.007330790627747774, |
| "step": 174 |
| }, |
| { |
| "epoch": 1.4, |
| "grad_norm": 0.050741445273160934, |
| "learning_rate": 2e-05, |
| "loss": 0.005321471951901913, |
| "step": 175 |
| }, |
| { |
| "epoch": 1.408, |
| "grad_norm": 0.06053789332509041, |
| "learning_rate": 2e-05, |
| "loss": 0.007614850997924805, |
| "step": 176 |
| }, |
| { |
| "epoch": 1.416, |
| "grad_norm": 0.045350395143032074, |
| "learning_rate": 2e-05, |
| "loss": 0.003982958849519491, |
| "step": 177 |
| }, |
| { |
| "epoch": 1.424, |
| "grad_norm": 0.04105447977781296, |
| "learning_rate": 2e-05, |
| "loss": 0.0032091487664729357, |
| "step": 178 |
| }, |
| { |
| "epoch": 1.432, |
| "grad_norm": 0.03485438972711563, |
| "learning_rate": 2e-05, |
| "loss": 0.0034181310329586267, |
| "step": 179 |
| }, |
| { |
| "epoch": 1.44, |
| "grad_norm": 0.0427364781498909, |
| "learning_rate": 2e-05, |
| "loss": 0.005363878328353167, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.448, |
| "grad_norm": 0.031005240976810455, |
| "learning_rate": 2e-05, |
| "loss": 0.003898187540471554, |
| "step": 181 |
| }, |
| { |
| "epoch": 1.456, |
| "grad_norm": 0.030742423608899117, |
| "learning_rate": 2e-05, |
| "loss": 0.0026354389265179634, |
| "step": 182 |
| }, |
| { |
| "epoch": 1.464, |
| "grad_norm": 0.04213688522577286, |
| "learning_rate": 2e-05, |
| "loss": 0.003235624171793461, |
| "step": 183 |
| }, |
| { |
| "epoch": 1.472, |
| "grad_norm": 0.06017177179455757, |
| "learning_rate": 2e-05, |
| "loss": 0.004921246785670519, |
| "step": 184 |
| }, |
| { |
| "epoch": 1.48, |
| "grad_norm": 0.05274427682161331, |
| "learning_rate": 2e-05, |
| "loss": 0.006431959569454193, |
| "step": 185 |
| }, |
| { |
| "epoch": 1.488, |
| "grad_norm": 0.03465743735432625, |
| "learning_rate": 2e-05, |
| "loss": 0.0017982947174459696, |
| "step": 186 |
| }, |
| { |
| "epoch": 1.496, |
| "grad_norm": 0.031407665461301804, |
| "learning_rate": 2e-05, |
| "loss": 0.002308888593688607, |
| "step": 187 |
| }, |
| { |
| "epoch": 1.504, |
| "grad_norm": 0.04751557856798172, |
| "learning_rate": 2e-05, |
| "loss": 0.0022526727989315987, |
| "step": 188 |
| }, |
| { |
| "epoch": 1.512, |
| "grad_norm": 0.03473098948597908, |
| "learning_rate": 2e-05, |
| "loss": 0.0023300114553421736, |
| "step": 189 |
| }, |
| { |
| "epoch": 1.52, |
| "grad_norm": 0.02745967172086239, |
| "learning_rate": 2e-05, |
| "loss": 0.002407602034509182, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.528, |
| "grad_norm": 0.05958602949976921, |
| "learning_rate": 2e-05, |
| "loss": 0.007107679732143879, |
| "step": 191 |
| }, |
| { |
| "epoch": 1.536, |
| "grad_norm": 0.05794493481516838, |
| "learning_rate": 2e-05, |
| "loss": 0.006767779588699341, |
| "step": 192 |
| }, |
| { |
| "epoch": 1.544, |
| "grad_norm": 0.037513379007577896, |
| "learning_rate": 2e-05, |
| "loss": 0.003455360187217593, |
| "step": 193 |
| }, |
| { |
| "epoch": 1.552, |
| "grad_norm": 0.03977490961551666, |
| "learning_rate": 2e-05, |
| "loss": 0.004349147900938988, |
| "step": 194 |
| }, |
| { |
| "epoch": 1.56, |
| "grad_norm": 0.03297000005841255, |
| "learning_rate": 2e-05, |
| "loss": 0.004090302158147097, |
| "step": 195 |
| }, |
| { |
| "epoch": 1.568, |
| "grad_norm": 0.03089851886034012, |
| "learning_rate": 2e-05, |
| "loss": 0.0032683194149285555, |
| "step": 196 |
| }, |
| { |
| "epoch": 1.576, |
| "grad_norm": 0.03224420174956322, |
| "learning_rate": 2e-05, |
| "loss": 0.0024244794622063637, |
| "step": 197 |
| }, |
| { |
| "epoch": 1.584, |
| "grad_norm": 0.030442802235484123, |
| "learning_rate": 2e-05, |
| "loss": 0.0014301001792773604, |
| "step": 198 |
| }, |
| { |
| "epoch": 1.592, |
| "grad_norm": 0.05126398056745529, |
| "learning_rate": 2e-05, |
| "loss": 0.0027298659551888704, |
| "step": 199 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 0.05423342436552048, |
| "learning_rate": 2e-05, |
| "loss": 0.006995318457484245, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.608, |
| "grad_norm": 0.02536446414887905, |
| "learning_rate": 2e-05, |
| "loss": 0.0029984633438289165, |
| "step": 201 |
| }, |
| { |
| "epoch": 1.616, |
| "grad_norm": 0.03072509542107582, |
| "learning_rate": 2e-05, |
| "loss": 0.0033400985412299633, |
| "step": 202 |
| }, |
| { |
| "epoch": 1.624, |
| "grad_norm": 0.055368877947330475, |
| "learning_rate": 2e-05, |
| "loss": 0.004085185006260872, |
| "step": 203 |
| }, |
| { |
| "epoch": 1.6320000000000001, |
| "grad_norm": 0.03939943388104439, |
| "learning_rate": 2e-05, |
| "loss": 0.003322000615298748, |
| "step": 204 |
| }, |
| { |
| "epoch": 1.6400000000000001, |
| "grad_norm": 0.044486552476882935, |
| "learning_rate": 2e-05, |
| "loss": 0.002586575224995613, |
| "step": 205 |
| }, |
| { |
| "epoch": 1.6480000000000001, |
| "grad_norm": 0.05059683322906494, |
| "learning_rate": 2e-05, |
| "loss": 0.010476172901690006, |
| "step": 206 |
| }, |
| { |
| "epoch": 1.6560000000000001, |
| "grad_norm": 0.04618003964424133, |
| "learning_rate": 2e-05, |
| "loss": 0.004687571432441473, |
| "step": 207 |
| }, |
| { |
| "epoch": 1.6640000000000001, |
| "grad_norm": 0.03308900445699692, |
| "learning_rate": 2e-05, |
| "loss": 0.004028619267046452, |
| "step": 208 |
| }, |
| { |
| "epoch": 1.6720000000000002, |
| "grad_norm": 0.038209252059459686, |
| "learning_rate": 2e-05, |
| "loss": 0.0027929157949984074, |
| "step": 209 |
| }, |
| { |
| "epoch": 1.6800000000000002, |
| "grad_norm": 0.034491200000047684, |
| "learning_rate": 2e-05, |
| "loss": 0.001778338453732431, |
| "step": 210 |
| }, |
| { |
| "epoch": 1.688, |
| "grad_norm": 0.07821262627840042, |
| "learning_rate": 2e-05, |
| "loss": 0.005375068634748459, |
| "step": 211 |
| }, |
| { |
| "epoch": 1.696, |
| "grad_norm": 0.04657049477100372, |
| "learning_rate": 2e-05, |
| "loss": 0.003136474872007966, |
| "step": 212 |
| }, |
| { |
| "epoch": 1.704, |
| "grad_norm": 0.05357731506228447, |
| "learning_rate": 2e-05, |
| "loss": 0.006239030975848436, |
| "step": 213 |
| }, |
| { |
| "epoch": 1.712, |
| "grad_norm": 0.04918161407113075, |
| "learning_rate": 2e-05, |
| "loss": 0.004385227337479591, |
| "step": 214 |
| }, |
| { |
| "epoch": 1.72, |
| "grad_norm": 0.04051123186945915, |
| "learning_rate": 2e-05, |
| "loss": 0.0024626629892736673, |
| "step": 215 |
| }, |
| { |
| "epoch": 1.728, |
| "grad_norm": 0.034734904766082764, |
| "learning_rate": 2e-05, |
| "loss": 0.0024731624871492386, |
| "step": 216 |
| }, |
| { |
| "epoch": 1.736, |
| "grad_norm": 0.06079041212797165, |
| "learning_rate": 2e-05, |
| "loss": 0.004821632523089647, |
| "step": 217 |
| }, |
| { |
| "epoch": 1.744, |
| "grad_norm": 0.04052048176527023, |
| "learning_rate": 2e-05, |
| "loss": 0.003583197481930256, |
| "step": 218 |
| }, |
| { |
| "epoch": 1.752, |
| "grad_norm": 0.03803209960460663, |
| "learning_rate": 2e-05, |
| "loss": 0.0032860925421118736, |
| "step": 219 |
| }, |
| { |
| "epoch": 1.76, |
| "grad_norm": 0.03718697279691696, |
| "learning_rate": 2e-05, |
| "loss": 0.0028506568633019924, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.768, |
| "grad_norm": 0.020328527316451073, |
| "learning_rate": 2e-05, |
| "loss": 0.0030127556528896093, |
| "step": 221 |
| }, |
| { |
| "epoch": 1.776, |
| "grad_norm": 0.03647777810692787, |
| "learning_rate": 2e-05, |
| "loss": 0.0029757029842585325, |
| "step": 222 |
| }, |
| { |
| "epoch": 1.784, |
| "grad_norm": 0.03103819489479065, |
| "learning_rate": 2e-05, |
| "loss": 0.002474813023582101, |
| "step": 223 |
| }, |
| { |
| "epoch": 1.792, |
| "grad_norm": 0.031560298055410385, |
| "learning_rate": 2e-05, |
| "loss": 0.002314931945875287, |
| "step": 224 |
| }, |
| { |
| "epoch": 1.8, |
| "grad_norm": 0.03549206256866455, |
| "learning_rate": 2e-05, |
| "loss": 0.0032754004932940006, |
| "step": 225 |
| }, |
| { |
| "epoch": 1.808, |
| "grad_norm": 0.03429936245083809, |
| "learning_rate": 2e-05, |
| "loss": 0.003921601455658674, |
| "step": 226 |
| }, |
| { |
| "epoch": 1.8159999999999998, |
| "grad_norm": 0.06855639815330505, |
| "learning_rate": 2e-05, |
| "loss": 0.006101913750171661, |
| "step": 227 |
| }, |
| { |
| "epoch": 1.8239999999999998, |
| "grad_norm": 0.08064403384923935, |
| "learning_rate": 2e-05, |
| "loss": 0.004588215611875057, |
| "step": 228 |
| }, |
| { |
| "epoch": 1.8319999999999999, |
| "grad_norm": 0.03363880142569542, |
| "learning_rate": 2e-05, |
| "loss": 0.001759383245371282, |
| "step": 229 |
| }, |
| { |
| "epoch": 1.8399999999999999, |
| "grad_norm": 0.04351300746202469, |
| "learning_rate": 2e-05, |
| "loss": 0.0032918578945100307, |
| "step": 230 |
| }, |
| { |
| "epoch": 1.8479999999999999, |
| "grad_norm": 0.024159353226423264, |
| "learning_rate": 2e-05, |
| "loss": 0.0019000337924808264, |
| "step": 231 |
| }, |
| { |
| "epoch": 1.8559999999999999, |
| "grad_norm": 0.04056349769234657, |
| "learning_rate": 2e-05, |
| "loss": 0.003138928906992078, |
| "step": 232 |
| }, |
| { |
| "epoch": 1.8639999999999999, |
| "grad_norm": 0.026023566722869873, |
| "learning_rate": 2e-05, |
| "loss": 0.0033341734670102596, |
| "step": 233 |
| }, |
| { |
| "epoch": 1.8719999999999999, |
| "grad_norm": 0.06581102311611176, |
| "learning_rate": 2e-05, |
| "loss": 0.0065653664059937, |
| "step": 234 |
| }, |
| { |
| "epoch": 1.88, |
| "grad_norm": 0.04903842508792877, |
| "learning_rate": 2e-05, |
| "loss": 0.005071278661489487, |
| "step": 235 |
| }, |
| { |
| "epoch": 1.888, |
| "grad_norm": 0.09484432637691498, |
| "learning_rate": 2e-05, |
| "loss": 0.0026917962823063135, |
| "step": 236 |
| }, |
| { |
| "epoch": 1.896, |
| "grad_norm": 0.029802817851305008, |
| "learning_rate": 2e-05, |
| "loss": 0.0020737922750413418, |
| "step": 237 |
| }, |
| { |
| "epoch": 1.904, |
| "grad_norm": 0.030020302161574364, |
| "learning_rate": 2e-05, |
| "loss": 0.001601328724063933, |
| "step": 238 |
| }, |
| { |
| "epoch": 1.912, |
| "grad_norm": 0.023855257779359818, |
| "learning_rate": 2e-05, |
| "loss": 0.002171756699681282, |
| "step": 239 |
| }, |
| { |
| "epoch": 1.92, |
| "grad_norm": 0.04743681848049164, |
| "learning_rate": 2e-05, |
| "loss": 0.005675522144883871, |
| "step": 240 |
| }, |
| { |
| "epoch": 1.928, |
| "grad_norm": 0.05006203055381775, |
| "learning_rate": 2e-05, |
| "loss": 0.005861939862370491, |
| "step": 241 |
| }, |
| { |
| "epoch": 1.936, |
| "grad_norm": 0.04128265008330345, |
| "learning_rate": 2e-05, |
| "loss": 0.005003426223993301, |
| "step": 242 |
| }, |
| { |
| "epoch": 1.944, |
| "grad_norm": 0.06207670643925667, |
| "learning_rate": 2e-05, |
| "loss": 0.004483950790017843, |
| "step": 243 |
| }, |
| { |
| "epoch": 1.952, |
| "grad_norm": 0.07824505865573883, |
| "learning_rate": 2e-05, |
| "loss": 0.004859112203121185, |
| "step": 244 |
| }, |
| { |
| "epoch": 1.96, |
| "grad_norm": 0.04579395800828934, |
| "learning_rate": 2e-05, |
| "loss": 0.004692354239523411, |
| "step": 245 |
| }, |
| { |
| "epoch": 1.968, |
| "grad_norm": 0.04180062189698219, |
| "learning_rate": 2e-05, |
| "loss": 0.00455811619758606, |
| "step": 246 |
| }, |
| { |
| "epoch": 1.976, |
| "grad_norm": 0.03964953497052193, |
| "learning_rate": 2e-05, |
| "loss": 0.0028610117733478546, |
| "step": 247 |
| }, |
| { |
| "epoch": 1.984, |
| "grad_norm": 0.03513772413134575, |
| "learning_rate": 2e-05, |
| "loss": 0.0025510897394269705, |
| "step": 248 |
| }, |
| { |
| "epoch": 1.992, |
| "grad_norm": 0.030972128733992577, |
| "learning_rate": 2e-05, |
| "loss": 0.00266153528355062, |
| "step": 249 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.06495940685272217, |
| "learning_rate": 2e-05, |
| "loss": 0.005316338501870632, |
| "step": 250 |
| }, |
| { |
| "epoch": 2.008, |
| "grad_norm": 0.03911746293306351, |
| "learning_rate": 2e-05, |
| "loss": 0.003454022342339158, |
| "step": 251 |
| }, |
| { |
| "epoch": 2.016, |
| "grad_norm": 0.03907843679189682, |
| "learning_rate": 2e-05, |
| "loss": 0.002617109566926956, |
| "step": 252 |
| }, |
| { |
| "epoch": 2.024, |
| "grad_norm": 0.0325784757733345, |
| "learning_rate": 2e-05, |
| "loss": 0.003027692437171936, |
| "step": 253 |
| }, |
| { |
| "epoch": 2.032, |
| "grad_norm": 0.04280916228890419, |
| "learning_rate": 2e-05, |
| "loss": 0.003725123591721058, |
| "step": 254 |
| }, |
| { |
| "epoch": 2.04, |
| "grad_norm": 0.03285227715969086, |
| "learning_rate": 2e-05, |
| "loss": 0.002418406307697296, |
| "step": 255 |
| }, |
| { |
| "epoch": 2.048, |
| "grad_norm": 0.04505843669176102, |
| "learning_rate": 2e-05, |
| "loss": 0.0030923697631806135, |
| "step": 256 |
| }, |
| { |
| "epoch": 2.056, |
| "grad_norm": 0.02693861722946167, |
| "learning_rate": 2e-05, |
| "loss": 0.0029823114164173603, |
| "step": 257 |
| }, |
| { |
| "epoch": 2.064, |
| "grad_norm": 0.03501451015472412, |
| "learning_rate": 2e-05, |
| "loss": 0.0022289499174803495, |
| "step": 258 |
| }, |
| { |
| "epoch": 2.072, |
| "grad_norm": 0.04407954588532448, |
| "learning_rate": 2e-05, |
| "loss": 0.005173118785023689, |
| "step": 259 |
| }, |
| { |
| "epoch": 2.08, |
| "grad_norm": 0.03353921324014664, |
| "learning_rate": 2e-05, |
| "loss": 0.003529382636770606, |
| "step": 260 |
| }, |
| { |
| "epoch": 2.088, |
| "grad_norm": 0.03029395453631878, |
| "learning_rate": 2e-05, |
| "loss": 0.0025696109514683485, |
| "step": 261 |
| }, |
| { |
| "epoch": 2.096, |
| "grad_norm": 0.043411292135715485, |
| "learning_rate": 2e-05, |
| "loss": 0.0032275894191116095, |
| "step": 262 |
| }, |
| { |
| "epoch": 2.104, |
| "grad_norm": 0.025156555697321892, |
| "learning_rate": 2e-05, |
| "loss": 0.0019941977225244045, |
| "step": 263 |
| }, |
| { |
| "epoch": 2.112, |
| "grad_norm": 0.05474105849862099, |
| "learning_rate": 2e-05, |
| "loss": 0.006130381021648645, |
| "step": 264 |
| }, |
| { |
| "epoch": 2.12, |
| "grad_norm": 0.026152821257710457, |
| "learning_rate": 2e-05, |
| "loss": 0.0037242607213556767, |
| "step": 265 |
| }, |
| { |
| "epoch": 2.128, |
| "grad_norm": 0.04658479988574982, |
| "learning_rate": 2e-05, |
| "loss": 0.0027500735595822334, |
| "step": 266 |
| }, |
| { |
| "epoch": 2.136, |
| "grad_norm": 0.04310692846775055, |
| "learning_rate": 2e-05, |
| "loss": 0.0021111357491463423, |
| "step": 267 |
| }, |
| { |
| "epoch": 2.144, |
| "grad_norm": 0.03860605135560036, |
| "learning_rate": 2e-05, |
| "loss": 0.004491395782679319, |
| "step": 268 |
| }, |
| { |
| "epoch": 2.152, |
| "grad_norm": 0.019513150677084923, |
| "learning_rate": 2e-05, |
| "loss": 0.0019580726511776447, |
| "step": 269 |
| }, |
| { |
| "epoch": 2.16, |
| "grad_norm": 0.03170303627848625, |
| "learning_rate": 2e-05, |
| "loss": 0.001965203322470188, |
| "step": 270 |
| }, |
| { |
| "epoch": 2.168, |
| "grad_norm": 0.02094845101237297, |
| "learning_rate": 2e-05, |
| "loss": 0.001287441118620336, |
| "step": 271 |
| }, |
| { |
| "epoch": 2.176, |
| "grad_norm": 0.03304935246706009, |
| "learning_rate": 2e-05, |
| "loss": 0.0032492834143340588, |
| "step": 272 |
| }, |
| { |
| "epoch": 2.184, |
| "grad_norm": 0.03120725229382515, |
| "learning_rate": 2e-05, |
| "loss": 0.0029026831034570932, |
| "step": 273 |
| }, |
| { |
| "epoch": 2.192, |
| "grad_norm": 0.03816477581858635, |
| "learning_rate": 2e-05, |
| "loss": 0.0024198577739298344, |
| "step": 274 |
| }, |
| { |
| "epoch": 2.2, |
| "grad_norm": 0.039687663316726685, |
| "learning_rate": 2e-05, |
| "loss": 0.0031529159750789404, |
| "step": 275 |
| }, |
| { |
| "epoch": 2.208, |
| "grad_norm": 0.026106711477041245, |
| "learning_rate": 2e-05, |
| "loss": 0.002224068157374859, |
| "step": 276 |
| }, |
| { |
| "epoch": 2.216, |
| "grad_norm": 0.027437372133135796, |
| "learning_rate": 2e-05, |
| "loss": 0.0014647672651335597, |
| "step": 277 |
| }, |
| { |
| "epoch": 2.224, |
| "grad_norm": 0.05910545587539673, |
| "learning_rate": 2e-05, |
| "loss": 0.00243214494548738, |
| "step": 278 |
| }, |
| { |
| "epoch": 2.232, |
| "grad_norm": 0.02404620312154293, |
| "learning_rate": 2e-05, |
| "loss": 0.00222192844375968, |
| "step": 279 |
| }, |
| { |
| "epoch": 2.24, |
| "grad_norm": 0.03369680047035217, |
| "learning_rate": 2e-05, |
| "loss": 0.001881018397398293, |
| "step": 280 |
| }, |
| { |
| "epoch": 2.248, |
| "grad_norm": 0.039129581302404404, |
| "learning_rate": 2e-05, |
| "loss": 0.0031921733170747757, |
| "step": 281 |
| }, |
| { |
| "epoch": 2.2560000000000002, |
| "grad_norm": 0.026752755045890808, |
| "learning_rate": 2e-05, |
| "loss": 0.0027538840658962727, |
| "step": 282 |
| }, |
| { |
| "epoch": 2.2640000000000002, |
| "grad_norm": 0.07003916054964066, |
| "learning_rate": 2e-05, |
| "loss": 0.005947021301835775, |
| "step": 283 |
| }, |
| { |
| "epoch": 2.2720000000000002, |
| "grad_norm": 0.031368181109428406, |
| "learning_rate": 2e-05, |
| "loss": 0.0035742626059800386, |
| "step": 284 |
| }, |
| { |
| "epoch": 2.2800000000000002, |
| "grad_norm": 0.027074171230196953, |
| "learning_rate": 2e-05, |
| "loss": 0.0021643084473907948, |
| "step": 285 |
| }, |
| { |
| "epoch": 2.288, |
| "grad_norm": 0.023025641217827797, |
| "learning_rate": 2e-05, |
| "loss": 0.0017899582162499428, |
| "step": 286 |
| }, |
| { |
| "epoch": 2.296, |
| "grad_norm": 0.03725794330239296, |
| "learning_rate": 2e-05, |
| "loss": 0.0034590037539601326, |
| "step": 287 |
| }, |
| { |
| "epoch": 2.304, |
| "grad_norm": 0.04669518023729324, |
| "learning_rate": 2e-05, |
| "loss": 0.003809502115473151, |
| "step": 288 |
| }, |
| { |
| "epoch": 2.312, |
| "grad_norm": 0.031050648540258408, |
| "learning_rate": 2e-05, |
| "loss": 0.003041312098503113, |
| "step": 289 |
| }, |
| { |
| "epoch": 2.32, |
| "grad_norm": 0.03626548871397972, |
| "learning_rate": 2e-05, |
| "loss": 0.0024002999998629093, |
| "step": 290 |
| }, |
| { |
| "epoch": 2.328, |
| "grad_norm": 0.06315096467733383, |
| "learning_rate": 2e-05, |
| "loss": 0.006624164525419474, |
| "step": 291 |
| }, |
| { |
| "epoch": 2.336, |
| "grad_norm": 0.04203896224498749, |
| "learning_rate": 2e-05, |
| "loss": 0.0019566724076867104, |
| "step": 292 |
| }, |
| { |
| "epoch": 2.344, |
| "grad_norm": 0.04634423181414604, |
| "learning_rate": 2e-05, |
| "loss": 0.0032502529211342335, |
| "step": 293 |
| }, |
| { |
| "epoch": 2.352, |
| "grad_norm": 0.04237474128603935, |
| "learning_rate": 2e-05, |
| "loss": 0.0031427519861608744, |
| "step": 294 |
| }, |
| { |
| "epoch": 2.36, |
| "grad_norm": 0.04034377261996269, |
| "learning_rate": 2e-05, |
| "loss": 0.00422636279836297, |
| "step": 295 |
| }, |
| { |
| "epoch": 2.368, |
| "grad_norm": 0.044460564851760864, |
| "learning_rate": 2e-05, |
| "loss": 0.0028524911031126976, |
| "step": 296 |
| }, |
| { |
| "epoch": 2.376, |
| "grad_norm": 0.0697890892624855, |
| "learning_rate": 2e-05, |
| "loss": 0.0036988966166973114, |
| "step": 297 |
| }, |
| { |
| "epoch": 2.384, |
| "grad_norm": 0.04324623942375183, |
| "learning_rate": 2e-05, |
| "loss": 0.0037890365347266197, |
| "step": 298 |
| }, |
| { |
| "epoch": 2.392, |
| "grad_norm": 0.02138633094727993, |
| "learning_rate": 2e-05, |
| "loss": 0.0021952460519969463, |
| "step": 299 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 0.03607005998492241, |
| "learning_rate": 2e-05, |
| "loss": 0.002349622081965208, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.408, |
| "grad_norm": 0.039072196930646896, |
| "learning_rate": 2e-05, |
| "loss": 0.0035770011600106955, |
| "step": 301 |
| }, |
| { |
| "epoch": 2.416, |
| "grad_norm": 0.02861890196800232, |
| "learning_rate": 2e-05, |
| "loss": 0.002283460693433881, |
| "step": 302 |
| }, |
| { |
| "epoch": 2.424, |
| "grad_norm": 0.035563308745622635, |
| "learning_rate": 2e-05, |
| "loss": 0.0026536290533840656, |
| "step": 303 |
| }, |
| { |
| "epoch": 2.432, |
| "grad_norm": 0.03916006535291672, |
| "learning_rate": 2e-05, |
| "loss": 0.003950632177293301, |
| "step": 304 |
| }, |
| { |
| "epoch": 2.44, |
| "grad_norm": 0.039335884153842926, |
| "learning_rate": 2e-05, |
| "loss": 0.003372431267052889, |
| "step": 305 |
| }, |
| { |
| "epoch": 2.448, |
| "grad_norm": 0.045527540147304535, |
| "learning_rate": 2e-05, |
| "loss": 0.0030923227313905954, |
| "step": 306 |
| }, |
| { |
| "epoch": 2.456, |
| "grad_norm": 0.04456671327352524, |
| "learning_rate": 2e-05, |
| "loss": 0.004458888433873653, |
| "step": 307 |
| }, |
| { |
| "epoch": 2.464, |
| "grad_norm": 0.03560574725270271, |
| "learning_rate": 2e-05, |
| "loss": 0.003171511460095644, |
| "step": 308 |
| }, |
| { |
| "epoch": 2.472, |
| "grad_norm": 0.036603160202503204, |
| "learning_rate": 2e-05, |
| "loss": 0.0028585128020495176, |
| "step": 309 |
| }, |
| { |
| "epoch": 2.48, |
| "grad_norm": 0.03335639089345932, |
| "learning_rate": 2e-05, |
| "loss": 0.0023816449102014303, |
| "step": 310 |
| }, |
| { |
| "epoch": 2.488, |
| "grad_norm": 0.030154302716255188, |
| "learning_rate": 2e-05, |
| "loss": 0.00486906711012125, |
| "step": 311 |
| }, |
| { |
| "epoch": 2.496, |
| "grad_norm": 0.040795303881168365, |
| "learning_rate": 2e-05, |
| "loss": 0.003139256499707699, |
| "step": 312 |
| }, |
| { |
| "epoch": 2.504, |
| "grad_norm": 0.04953761771321297, |
| "learning_rate": 2e-05, |
| "loss": 0.0039007822051644325, |
| "step": 313 |
| }, |
| { |
| "epoch": 2.512, |
| "grad_norm": 0.12026029080152512, |
| "learning_rate": 2e-05, |
| "loss": 0.0038295963313430548, |
| "step": 314 |
| }, |
| { |
| "epoch": 2.52, |
| "grad_norm": 0.057940348982810974, |
| "learning_rate": 2e-05, |
| "loss": 0.0033367923460900784, |
| "step": 315 |
| }, |
| { |
| "epoch": 2.528, |
| "grad_norm": 0.028937026858329773, |
| "learning_rate": 2e-05, |
| "loss": 0.001545312232337892, |
| "step": 316 |
| }, |
| { |
| "epoch": 2.536, |
| "grad_norm": 0.03167141228914261, |
| "learning_rate": 2e-05, |
| "loss": 0.0028309039771556854, |
| "step": 317 |
| }, |
| { |
| "epoch": 2.544, |
| "grad_norm": 0.04004928842186928, |
| "learning_rate": 2e-05, |
| "loss": 0.002698655240237713, |
| "step": 318 |
| }, |
| { |
| "epoch": 2.552, |
| "grad_norm": 0.02045566588640213, |
| "learning_rate": 2e-05, |
| "loss": 0.0012601229827851057, |
| "step": 319 |
| }, |
| { |
| "epoch": 2.56, |
| "grad_norm": 0.03342553600668907, |
| "learning_rate": 2e-05, |
| "loss": 0.001859499723650515, |
| "step": 320 |
| }, |
| { |
| "epoch": 2.568, |
| "grad_norm": 0.025396211072802544, |
| "learning_rate": 2e-05, |
| "loss": 0.0020996080711483955, |
| "step": 321 |
| }, |
| { |
| "epoch": 2.576, |
| "grad_norm": 0.03771315887570381, |
| "learning_rate": 2e-05, |
| "loss": 0.005717152263969183, |
| "step": 322 |
| }, |
| { |
| "epoch": 2.584, |
| "grad_norm": 0.02477932535111904, |
| "learning_rate": 2e-05, |
| "loss": 0.0024716444313526154, |
| "step": 323 |
| }, |
| { |
| "epoch": 2.592, |
| "grad_norm": 0.031247377395629883, |
| "learning_rate": 2e-05, |
| "loss": 0.00390324997715652, |
| "step": 324 |
| }, |
| { |
| "epoch": 2.6, |
| "grad_norm": 0.04134252667427063, |
| "learning_rate": 2e-05, |
| "loss": 0.003920188173651695, |
| "step": 325 |
| }, |
| { |
| "epoch": 2.608, |
| "grad_norm": 0.02774622291326523, |
| "learning_rate": 2e-05, |
| "loss": 0.0022154548205435276, |
| "step": 326 |
| }, |
| { |
| "epoch": 2.616, |
| "grad_norm": 0.03957367688417435, |
| "learning_rate": 2e-05, |
| "loss": 0.0023058855440467596, |
| "step": 327 |
| }, |
| { |
| "epoch": 2.624, |
| "grad_norm": 0.04483708739280701, |
| "learning_rate": 2e-05, |
| "loss": 0.002552008954808116, |
| "step": 328 |
| }, |
| { |
| "epoch": 2.632, |
| "grad_norm": 0.02093776874244213, |
| "learning_rate": 2e-05, |
| "loss": 0.0027019393164664507, |
| "step": 329 |
| }, |
| { |
| "epoch": 2.64, |
| "grad_norm": 0.06742503494024277, |
| "learning_rate": 2e-05, |
| "loss": 0.008755444549024105, |
| "step": 330 |
| }, |
| { |
| "epoch": 2.648, |
| "grad_norm": 0.038483329117298126, |
| "learning_rate": 2e-05, |
| "loss": 0.0028858440928161144, |
| "step": 331 |
| }, |
| { |
| "epoch": 2.656, |
| "grad_norm": 0.0329829640686512, |
| "learning_rate": 2e-05, |
| "loss": 0.0024300175718963146, |
| "step": 332 |
| }, |
| { |
| "epoch": 2.664, |
| "grad_norm": 0.051471542567014694, |
| "learning_rate": 2e-05, |
| "loss": 0.003998158499598503, |
| "step": 333 |
| }, |
| { |
| "epoch": 2.672, |
| "grad_norm": 0.04422187805175781, |
| "learning_rate": 2e-05, |
| "loss": 0.004651157651096582, |
| "step": 334 |
| }, |
| { |
| "epoch": 2.68, |
| "grad_norm": 0.02528465911746025, |
| "learning_rate": 2e-05, |
| "loss": 0.0021933559328317642, |
| "step": 335 |
| }, |
| { |
| "epoch": 2.6879999999999997, |
| "grad_norm": 0.030596323311328888, |
| "learning_rate": 2e-05, |
| "loss": 0.003947809804230928, |
| "step": 336 |
| }, |
| { |
| "epoch": 2.6959999999999997, |
| "grad_norm": 0.04210244491696358, |
| "learning_rate": 2e-05, |
| "loss": 0.003154753241688013, |
| "step": 337 |
| }, |
| { |
| "epoch": 2.7039999999999997, |
| "grad_norm": 0.06121843308210373, |
| "learning_rate": 2e-05, |
| "loss": 0.0029088810551911592, |
| "step": 338 |
| }, |
| { |
| "epoch": 2.7119999999999997, |
| "grad_norm": 0.03322814777493477, |
| "learning_rate": 2e-05, |
| "loss": 0.0026830481365323067, |
| "step": 339 |
| }, |
| { |
| "epoch": 2.7199999999999998, |
| "grad_norm": 0.03890882432460785, |
| "learning_rate": 2e-05, |
| "loss": 0.002993557136505842, |
| "step": 340 |
| }, |
| { |
| "epoch": 2.7279999999999998, |
| "grad_norm": 0.03936028108000755, |
| "learning_rate": 2e-05, |
| "loss": 0.0023103575222194195, |
| "step": 341 |
| }, |
| { |
| "epoch": 2.7359999999999998, |
| "grad_norm": 0.03645415976643562, |
| "learning_rate": 2e-05, |
| "loss": 0.0036545468028634787, |
| "step": 342 |
| }, |
| { |
| "epoch": 2.7439999999999998, |
| "grad_norm": 0.04062900319695473, |
| "learning_rate": 2e-05, |
| "loss": 0.0036479239352047443, |
| "step": 343 |
| }, |
| { |
| "epoch": 2.752, |
| "grad_norm": 0.049033768475055695, |
| "learning_rate": 2e-05, |
| "loss": 0.0034971858840435743, |
| "step": 344 |
| }, |
| { |
| "epoch": 2.76, |
| "grad_norm": 0.03926759585738182, |
| "learning_rate": 2e-05, |
| "loss": 0.0031690222676843405, |
| "step": 345 |
| }, |
| { |
| "epoch": 2.768, |
| "grad_norm": 0.06413634121417999, |
| "learning_rate": 2e-05, |
| "loss": 0.004980374127626419, |
| "step": 346 |
| }, |
| { |
| "epoch": 2.776, |
| "grad_norm": 0.04801516979932785, |
| "learning_rate": 2e-05, |
| "loss": 0.0037698009982705116, |
| "step": 347 |
| }, |
| { |
| "epoch": 2.784, |
| "grad_norm": 0.03702208027243614, |
| "learning_rate": 2e-05, |
| "loss": 0.001877334201708436, |
| "step": 348 |
| }, |
| { |
| "epoch": 2.792, |
| "grad_norm": 0.03120242804288864, |
| "learning_rate": 2e-05, |
| "loss": 0.0022658274974673986, |
| "step": 349 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 0.048766590654850006, |
| "learning_rate": 2e-05, |
| "loss": 0.00202578236348927, |
| "step": 350 |
| }, |
| { |
| "epoch": 2.808, |
| "grad_norm": 0.029100535437464714, |
| "learning_rate": 2e-05, |
| "loss": 0.004179732408374548, |
| "step": 351 |
| }, |
| { |
| "epoch": 2.816, |
| "grad_norm": 0.04512881860136986, |
| "learning_rate": 2e-05, |
| "loss": 0.002607448725029826, |
| "step": 352 |
| }, |
| { |
| "epoch": 2.824, |
| "grad_norm": 0.07799641042947769, |
| "learning_rate": 2e-05, |
| "loss": 0.004347057081758976, |
| "step": 353 |
| }, |
| { |
| "epoch": 2.832, |
| "grad_norm": 0.04792260378599167, |
| "learning_rate": 2e-05, |
| "loss": 0.003629413666203618, |
| "step": 354 |
| }, |
| { |
| "epoch": 2.84, |
| "grad_norm": 0.11515104025602341, |
| "learning_rate": 2e-05, |
| "loss": 0.006446151994168758, |
| "step": 355 |
| }, |
| { |
| "epoch": 2.848, |
| "grad_norm": 0.023408330976963043, |
| "learning_rate": 2e-05, |
| "loss": 0.0017018368234857917, |
| "step": 356 |
| }, |
| { |
| "epoch": 2.856, |
| "grad_norm": 0.03008626215159893, |
| "learning_rate": 2e-05, |
| "loss": 0.002287666779011488, |
| "step": 357 |
| }, |
| { |
| "epoch": 2.864, |
| "grad_norm": 0.03706612437963486, |
| "learning_rate": 2e-05, |
| "loss": 0.0028982325457036495, |
| "step": 358 |
| }, |
| { |
| "epoch": 2.872, |
| "grad_norm": 0.020318036898970604, |
| "learning_rate": 2e-05, |
| "loss": 0.0018703520763665438, |
| "step": 359 |
| }, |
| { |
| "epoch": 2.88, |
| "grad_norm": 0.06123220548033714, |
| "learning_rate": 2e-05, |
| "loss": 0.0029717511497437954, |
| "step": 360 |
| }, |
| { |
| "epoch": 2.888, |
| "grad_norm": 0.055422086268663406, |
| "learning_rate": 2e-05, |
| "loss": 0.004031006712466478, |
| "step": 361 |
| }, |
| { |
| "epoch": 2.896, |
| "grad_norm": 0.036786146461963654, |
| "learning_rate": 2e-05, |
| "loss": 0.003937885165214539, |
| "step": 362 |
| }, |
| { |
| "epoch": 2.904, |
| "grad_norm": 0.038341671228408813, |
| "learning_rate": 2e-05, |
| "loss": 0.002436596667394042, |
| "step": 363 |
| }, |
| { |
| "epoch": 2.912, |
| "grad_norm": 0.03609547019004822, |
| "learning_rate": 2e-05, |
| "loss": 0.004547444172203541, |
| "step": 364 |
| }, |
| { |
| "epoch": 2.92, |
| "grad_norm": 0.06485926359891891, |
| "learning_rate": 2e-05, |
| "loss": 0.003946429584175348, |
| "step": 365 |
| }, |
| { |
| "epoch": 2.928, |
| "grad_norm": 0.04949840530753136, |
| "learning_rate": 2e-05, |
| "loss": 0.00502239353954792, |
| "step": 366 |
| }, |
| { |
| "epoch": 2.936, |
| "grad_norm": 0.0354098416864872, |
| "learning_rate": 2e-05, |
| "loss": 0.002601196290925145, |
| "step": 367 |
| }, |
| { |
| "epoch": 2.944, |
| "grad_norm": 0.025988208130002022, |
| "learning_rate": 2e-05, |
| "loss": 0.0036114591639488935, |
| "step": 368 |
| }, |
| { |
| "epoch": 2.952, |
| "grad_norm": 0.034603431820869446, |
| "learning_rate": 2e-05, |
| "loss": 0.0026605837047100067, |
| "step": 369 |
| }, |
| { |
| "epoch": 2.96, |
| "grad_norm": 0.045072562992572784, |
| "learning_rate": 2e-05, |
| "loss": 0.002992109628394246, |
| "step": 370 |
| }, |
| { |
| "epoch": 2.968, |
| "grad_norm": 0.07029342651367188, |
| "learning_rate": 2e-05, |
| "loss": 0.004379983991384506, |
| "step": 371 |
| }, |
| { |
| "epoch": 2.976, |
| "grad_norm": 0.013431387022137642, |
| "learning_rate": 2e-05, |
| "loss": 0.0030165393836796284, |
| "step": 372 |
| }, |
| { |
| "epoch": 2.984, |
| "grad_norm": 0.046111706644296646, |
| "learning_rate": 2e-05, |
| "loss": 0.0029378868639469147, |
| "step": 373 |
| }, |
| { |
| "epoch": 2.992, |
| "grad_norm": 0.05583514645695686, |
| "learning_rate": 2e-05, |
| "loss": 0.0040816557593643665, |
| "step": 374 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 0.03255736827850342, |
| "learning_rate": 2e-05, |
| "loss": 0.0035074178595095873, |
| "step": 375 |
| }, |
| { |
| "epoch": 3.008, |
| "grad_norm": 0.03124692291021347, |
| "learning_rate": 2e-05, |
| "loss": 0.0031269516330212355, |
| "step": 376 |
| }, |
| { |
| "epoch": 3.016, |
| "grad_norm": 0.05301191285252571, |
| "learning_rate": 2e-05, |
| "loss": 0.004872915334999561, |
| "step": 377 |
| }, |
| { |
| "epoch": 3.024, |
| "grad_norm": 0.037463486194610596, |
| "learning_rate": 2e-05, |
| "loss": 0.0030003672000020742, |
| "step": 378 |
| }, |
| { |
| "epoch": 3.032, |
| "grad_norm": 0.03362922742962837, |
| "learning_rate": 2e-05, |
| "loss": 0.0018557708244770765, |
| "step": 379 |
| }, |
| { |
| "epoch": 3.04, |
| "grad_norm": 0.042213547974824905, |
| "learning_rate": 2e-05, |
| "loss": 0.0038657563272863626, |
| "step": 380 |
| }, |
| { |
| "epoch": 3.048, |
| "grad_norm": 0.020529089495539665, |
| "learning_rate": 2e-05, |
| "loss": 0.0015072536189109087, |
| "step": 381 |
| }, |
| { |
| "epoch": 3.056, |
| "grad_norm": 0.07512009888887405, |
| "learning_rate": 2e-05, |
| "loss": 0.004759274888783693, |
| "step": 382 |
| }, |
| { |
| "epoch": 3.064, |
| "grad_norm": 0.03783705085515976, |
| "learning_rate": 2e-05, |
| "loss": 0.0032726821955293417, |
| "step": 383 |
| }, |
| { |
| "epoch": 3.072, |
| "grad_norm": 0.04495864734053612, |
| "learning_rate": 2e-05, |
| "loss": 0.003068119753152132, |
| "step": 384 |
| }, |
| { |
| "epoch": 3.08, |
| "grad_norm": 0.03879435732960701, |
| "learning_rate": 2e-05, |
| "loss": 0.0027257727924734354, |
| "step": 385 |
| }, |
| { |
| "epoch": 3.088, |
| "grad_norm": 0.049851685762405396, |
| "learning_rate": 2e-05, |
| "loss": 0.004501521587371826, |
| "step": 386 |
| }, |
| { |
| "epoch": 3.096, |
| "grad_norm": 0.037314675748348236, |
| "learning_rate": 2e-05, |
| "loss": 0.0031016827560961246, |
| "step": 387 |
| }, |
| { |
| "epoch": 3.104, |
| "grad_norm": 0.032761648297309875, |
| "learning_rate": 2e-05, |
| "loss": 0.0029233950190246105, |
| "step": 388 |
| }, |
| { |
| "epoch": 3.112, |
| "grad_norm": 0.0324791744351387, |
| "learning_rate": 2e-05, |
| "loss": 0.0026512390468269587, |
| "step": 389 |
| }, |
| { |
| "epoch": 3.12, |
| "grad_norm": 0.03452916815876961, |
| "learning_rate": 2e-05, |
| "loss": 0.003484592540189624, |
| "step": 390 |
| }, |
| { |
| "epoch": 3.128, |
| "grad_norm": 0.04120925813913345, |
| "learning_rate": 2e-05, |
| "loss": 0.0027602710761129856, |
| "step": 391 |
| }, |
| { |
| "epoch": 3.136, |
| "grad_norm": 0.03730145841836929, |
| "learning_rate": 2e-05, |
| "loss": 0.003531615249812603, |
| "step": 392 |
| }, |
| { |
| "epoch": 3.144, |
| "grad_norm": 0.03210754320025444, |
| "learning_rate": 2e-05, |
| "loss": 0.0023888859432190657, |
| "step": 393 |
| }, |
| { |
| "epoch": 3.152, |
| "grad_norm": 0.036783989518880844, |
| "learning_rate": 2e-05, |
| "loss": 0.0027839350514113903, |
| "step": 394 |
| }, |
| { |
| "epoch": 3.16, |
| "grad_norm": 0.03480631858110428, |
| "learning_rate": 2e-05, |
| "loss": 0.0027098648715764284, |
| "step": 395 |
| }, |
| { |
| "epoch": 3.168, |
| "grad_norm": 0.03560803085565567, |
| "learning_rate": 2e-05, |
| "loss": 0.003286743303760886, |
| "step": 396 |
| }, |
| { |
| "epoch": 3.176, |
| "grad_norm": 0.02866840735077858, |
| "learning_rate": 2e-05, |
| "loss": 0.0024349656887352467, |
| "step": 397 |
| }, |
| { |
| "epoch": 3.184, |
| "grad_norm": 0.025658991187810898, |
| "learning_rate": 2e-05, |
| "loss": 0.0043978262692689896, |
| "step": 398 |
| }, |
| { |
| "epoch": 3.192, |
| "grad_norm": 0.030603215098381042, |
| "learning_rate": 2e-05, |
| "loss": 0.0023162788711488247, |
| "step": 399 |
| }, |
| { |
| "epoch": 3.2, |
| "grad_norm": 0.05031350627541542, |
| "learning_rate": 2e-05, |
| "loss": 0.0037095744628459215, |
| "step": 400 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 400, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "save_steps": 200, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
