| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.0, |
| "eval_steps": 500, |
| "global_step": 1326, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "cos_sim": -0.0038788579404354095, |
| "distill_lambda": 0.05000000447034836, |
| "distill_loss": 1.0038789510726929, |
| "epoch": 0.0015082956259426848, |
| "grad_norm": 58.99883270263672, |
| "learning_rate": 2.2388059701492537e-07, |
| "loss": 2.520024538040161, |
| "sft_loss": 0.02855682373046875, |
| "step": 1, |
| "token_acc": 0.8815537084398977 |
| }, |
| { |
| "cos_sim": 0.00022824517347746424, |
| "distill_lambda": 0.050000001986821495, |
| "distill_loss": 0.9997717539469401, |
| "epoch": 0.015082956259426848, |
| "grad_norm": 45.729373931884766, |
| "learning_rate": 2.2388059701492535e-06, |
| "loss": 2.368698332044813, |
| "sft_loss": 0.024033228556315105, |
| "step": 10, |
| "token_acc": 0.8868504628691127 |
| }, |
| { |
| "cos_sim": 0.09473565220832825, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.9052643775939941, |
| "epoch": 0.030165912518853696, |
| "grad_norm": 25.559162139892578, |
| "learning_rate": 4.477611940298507e-06, |
| "loss": 1.6844642639160157, |
| "sft_loss": 0.007376289367675782, |
| "step": 20, |
| "token_acc": 0.9129723341716918 |
| }, |
| { |
| "cos_sim": 0.46269831657409666, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.5373017311096191, |
| "epoch": 0.04524886877828054, |
| "grad_norm": 11.83865737915039, |
| "learning_rate": 6.716417910447762e-06, |
| "loss": 1.0358287811279296, |
| "sft_loss": 0.0055045604705810545, |
| "step": 30, |
| "token_acc": 0.9293570753230381 |
| }, |
| { |
| "cos_sim": 0.7524478912353516, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.24755215644836426, |
| "epoch": 0.06033182503770739, |
| "grad_norm": 8.810158729553223, |
| "learning_rate": 8.955223880597014e-06, |
| "loss": 0.5503463745117188, |
| "sft_loss": 0.004820716381072998, |
| "step": 40, |
| "token_acc": 0.936442432636975 |
| }, |
| { |
| "cos_sim": 0.8826584815979004, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.1173414945602417, |
| "epoch": 0.07541478129713423, |
| "grad_norm": 6.174355506896973, |
| "learning_rate": 1.119402985074627e-05, |
| "loss": 0.3322910308837891, |
| "sft_loss": 0.004517018795013428, |
| "step": 50, |
| "token_acc": 0.9420423475551599 |
| }, |
| { |
| "cos_sim": 0.9243446350097656, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0756553053855896, |
| "epoch": 0.09049773755656108, |
| "grad_norm": 4.474364757537842, |
| "learning_rate": 1.3432835820895523e-05, |
| "loss": 0.26489078998565674, |
| "sft_loss": 0.004495072364807129, |
| "step": 60, |
| "token_acc": 0.941977054343285 |
| }, |
| { |
| "cos_sim": 0.9318131446838379, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.06818689107894897, |
| "epoch": 0.10558069381598793, |
| "grad_norm": 2.775825023651123, |
| "learning_rate": 1.4999789854529206e-05, |
| "loss": 0.2519868850708008, |
| "sft_loss": 0.004465246200561523, |
| "step": 70, |
| "token_acc": 0.9433893684688778 |
| }, |
| { |
| "cos_sim": 0.9455140113830567, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.054486030340194704, |
| "epoch": 0.12066365007541478, |
| "grad_norm": 4.228744029998779, |
| "learning_rate": 1.4996054262643387e-05, |
| "loss": 0.22749404907226561, |
| "sft_loss": 0.004384887218475342, |
| "step": 80, |
| "token_acc": 0.9445034140777238 |
| }, |
| { |
| "cos_sim": 0.9479646682739258, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.052035284042358396, |
| "epoch": 0.13574660633484162, |
| "grad_norm": 6.396712779998779, |
| "learning_rate": 1.498765144861652e-05, |
| "loss": 0.22304110527038573, |
| "sft_loss": 0.0043682694435119625, |
| "step": 90, |
| "token_acc": 0.9444644410450224 |
| }, |
| { |
| "cos_sim": 0.9392847061157227, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.060715246200561526, |
| "epoch": 0.15082956259426847, |
| "grad_norm": 1.407213568687439, |
| "learning_rate": 1.4974586644241e-05, |
| "loss": 0.23490571975708008, |
| "sft_loss": 0.004305040836334229, |
| "step": 100, |
| "token_acc": 0.9454039405258917 |
| }, |
| { |
| "cos_sim": 0.9490450859069824, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.05095493793487549, |
| "epoch": 0.16591251885369532, |
| "grad_norm": 2.289855480194092, |
| "learning_rate": 1.4956867983975587e-05, |
| "loss": 0.21652469635009766, |
| "sft_loss": 0.004218649864196777, |
| "step": 110, |
| "token_acc": 0.947564196739759 |
| }, |
| { |
| "cos_sim": 0.9502164840698242, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04978338181972504, |
| "epoch": 0.18099547511312217, |
| "grad_norm": 1.5713039636611938, |
| "learning_rate": 1.493450649988069e-05, |
| "loss": 0.2129663944244385, |
| "sft_loss": 0.004166030883789062, |
| "step": 120, |
| "token_acc": 0.948197482197355 |
| }, |
| { |
| "cos_sim": 0.9450735092163086, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.054926556348800656, |
| "epoch": 0.19607843137254902, |
| "grad_norm": 1.064908504486084, |
| "learning_rate": 1.4907516114749562e-05, |
| "loss": 0.22033605575561524, |
| "sft_loss": 0.004139173030853272, |
| "step": 130, |
| "token_acc": 0.9483397101983743 |
| }, |
| { |
| "cos_sim": 0.9489644050598145, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.05103568434715271, |
| "epoch": 0.21116138763197587, |
| "grad_norm": 1.0259686708450317, |
| "learning_rate": 1.487591363343962e-05, |
| "loss": 0.2132336139678955, |
| "sft_loss": 0.004111766815185547, |
| "step": 140, |
| "token_acc": 0.9494511370894669 |
| }, |
| { |
| "cos_sim": 0.9520336151123047, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04796643853187561, |
| "epoch": 0.22624434389140272, |
| "grad_norm": 1.421288013458252, |
| "learning_rate": 1.4839718732409347e-05, |
| "loss": 0.20410797595977784, |
| "sft_loss": 0.0039800524711608885, |
| "step": 150, |
| "token_acc": 0.9502197768419759 |
| }, |
| { |
| "cos_sim": 0.949534797668457, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.050465166568756104, |
| "epoch": 0.24132730015082957, |
| "grad_norm": 1.156361699104309, |
| "learning_rate": 1.4798953947467258e-05, |
| "loss": 0.21003410816192628, |
| "sft_loss": 0.004040306806564331, |
| "step": 160, |
| "token_acc": 0.9498282518162751 |
| }, |
| { |
| "cos_sim": 0.948972511291504, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.051027512550354, |
| "epoch": 0.2564102564102564, |
| "grad_norm": 1.0698307752609253, |
| "learning_rate": 1.4753644659740567e-05, |
| "loss": 0.20925650596618653, |
| "sft_loss": 0.003987890481948852, |
| "step": 170, |
| "token_acc": 0.9506242473208433 |
| }, |
| { |
| "cos_sim": 0.9544513702392579, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.045548611879348756, |
| "epoch": 0.27149321266968324, |
| "grad_norm": 1.2081609964370728, |
| "learning_rate": 1.4703819079872297e-05, |
| "loss": 0.19707401990890502, |
| "sft_loss": 0.0038811326026916503, |
| "step": 180, |
| "token_acc": 0.951166674188744 |
| }, |
| { |
| "cos_sim": 0.9517936706542969, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04820637702941895, |
| "epoch": 0.2865761689291101, |
| "grad_norm": 1.3792004585266113, |
| "learning_rate": 1.4649508230456655e-05, |
| "loss": 0.1991652250289917, |
| "sft_loss": 0.0038135945796966553, |
| "step": 190, |
| "token_acc": 0.9515554316167814 |
| }, |
| { |
| "cos_sim": 0.9557013511657715, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.044298702478408815, |
| "epoch": 0.30165912518853694, |
| "grad_norm": 1.2754738330841064, |
| "learning_rate": 1.459074592672363e-05, |
| "loss": 0.1925476908683777, |
| "sft_loss": 0.003802180290222168, |
| "step": 200, |
| "token_acc": 0.9524128472554461 |
| }, |
| { |
| "cos_sim": 0.9551361083984375, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04486394822597504, |
| "epoch": 0.3167420814479638, |
| "grad_norm": 1.6434708833694458, |
| "learning_rate": 1.4527568755484822e-05, |
| "loss": 0.18934518098831177, |
| "sft_loss": 0.003673839569091797, |
| "step": 210, |
| "token_acc": 0.9543265859473611 |
| }, |
| { |
| "cos_sim": 0.9531267166137696, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04687337875366211, |
| "epoch": 0.33182503770739064, |
| "grad_norm": 0.9669171571731567, |
| "learning_rate": 1.4460016052353615e-05, |
| "loss": 0.19388854503631592, |
| "sft_loss": 0.0037153482437133787, |
| "step": 220, |
| "token_acc": 0.9548988874546683 |
| }, |
| { |
| "cos_sim": 0.9562634468078614, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043736580014228824, |
| "epoch": 0.3469079939668175, |
| "grad_norm": 1.262874960899353, |
| "learning_rate": 1.43881298772539e-05, |
| "loss": 0.18832703828811645, |
| "sft_loss": 0.00369839072227478, |
| "step": 230, |
| "token_acc": 0.9528906529589746 |
| }, |
| { |
| "cos_sim": 0.9578750610351563, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04212484657764435, |
| "epoch": 0.36199095022624433, |
| "grad_norm": 0.8858125805854797, |
| "learning_rate": 1.4311954988232549e-05, |
| "loss": 0.18181778192520143, |
| "sft_loss": 0.003575563430786133, |
| "step": 240, |
| "token_acc": 0.9552487502975482 |
| }, |
| { |
| "cos_sim": 0.9587590217590332, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04124096930027008, |
| "epoch": 0.3770739064856712, |
| "grad_norm": 0.7233920693397522, |
| "learning_rate": 1.4231538813591998e-05, |
| "loss": 0.18377578258514404, |
| "sft_loss": 0.0036809444427490234, |
| "step": 250, |
| "token_acc": 0.954867890147078 |
| }, |
| { |
| "cos_sim": 0.957796859741211, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04220314919948578, |
| "epoch": 0.39215686274509803, |
| "grad_norm": 0.83754563331604, |
| "learning_rate": 1.4146931422360251e-05, |
| "loss": 0.18479570150375366, |
| "sft_loss": 0.003664708137512207, |
| "step": 260, |
| "token_acc": 0.954435550482827 |
| }, |
| { |
| "cos_sim": 0.9538604736328125, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0461395263671875, |
| "epoch": 0.4072398190045249, |
| "grad_norm": 0.7190311551094055, |
| "learning_rate": 1.4058185493116703e-05, |
| "loss": 0.19155261516571045, |
| "sft_loss": 0.0036790430545806884, |
| "step": 270, |
| "token_acc": 0.9548158618225874 |
| }, |
| { |
| "cos_sim": 0.9538644790649414, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04613546431064606, |
| "epoch": 0.42232277526395173, |
| "grad_norm": 0.8648778200149536, |
| "learning_rate": 1.396535628119321e-05, |
| "loss": 0.18670086860656737, |
| "sft_loss": 0.003527629375457764, |
| "step": 280, |
| "token_acc": 0.9557889084454521 |
| }, |
| { |
| "cos_sim": 0.9595279693603516, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04047197699546814, |
| "epoch": 0.4374057315233786, |
| "grad_norm": 0.8901738524436951, |
| "learning_rate": 1.3868501584270802e-05, |
| "loss": 0.1772393226623535, |
| "sft_loss": 0.003515130281448364, |
| "step": 290, |
| "token_acc": 0.9564196791374593 |
| }, |
| { |
| "cos_sim": 0.9556629180908203, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04433700740337372, |
| "epoch": 0.45248868778280543, |
| "grad_norm": 0.7970823645591736, |
| "learning_rate": 1.3767681706393457e-05, |
| "loss": 0.1882307529449463, |
| "sft_loss": 0.0036653608083724977, |
| "step": 300, |
| "token_acc": 0.9546085057403417 |
| }, |
| { |
| "cos_sim": 0.9546260833740234, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04537382423877716, |
| "epoch": 0.4675716440422323, |
| "grad_norm": 0.849625825881958, |
| "learning_rate": 1.3662959420421375e-05, |
| "loss": 0.1848926067352295, |
| "sft_loss": 0.0035092025995254517, |
| "step": 310, |
| "token_acc": 0.9562830728057904 |
| }, |
| { |
| "cos_sim": 0.9584080696105957, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0415919303894043, |
| "epoch": 0.48265460030165913, |
| "grad_norm": 1.1224209070205688, |
| "learning_rate": 1.35543999289471e-05, |
| "loss": 0.17938269376754762, |
| "sft_loss": 0.00352611243724823, |
| "step": 320, |
| "token_acc": 0.9556174728604272 |
| }, |
| { |
| "cos_sim": 0.9568581581115723, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04314180314540863, |
| "epoch": 0.497737556561086, |
| "grad_norm": 0.6940692067146301, |
| "learning_rate": 1.3442070823698822e-05, |
| "loss": 0.18326592445373535, |
| "sft_loss": 0.0035699695348739625, |
| "step": 330, |
| "token_acc": 0.9559384414466557 |
| }, |
| { |
| "cos_sim": 0.9564298629760742, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04357010424137116, |
| "epoch": 0.5128205128205128, |
| "grad_norm": 0.6807535290718079, |
| "learning_rate": 1.332604204345616e-05, |
| "loss": 0.18083639144897462, |
| "sft_loss": 0.0034726321697235107, |
| "step": 340, |
| "token_acc": 0.957401203499357 |
| }, |
| { |
| "cos_sim": 0.9551042556762696, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04489580094814301, |
| "epoch": 0.5279034690799397, |
| "grad_norm": 0.9952406883239746, |
| "learning_rate": 1.3206385830504614e-05, |
| "loss": 0.18671003580093384, |
| "sft_loss": 0.0035898983478546143, |
| "step": 350, |
| "token_acc": 0.9559120709602776 |
| }, |
| { |
| "cos_sim": 0.9583676338195801, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04163242280483246, |
| "epoch": 0.5429864253393665, |
| "grad_norm": 0.7764410972595215, |
| "learning_rate": 1.3083176685655786e-05, |
| "loss": 0.17547743320465087, |
| "sft_loss": 0.0034020483493804933, |
| "step": 360, |
| "token_acc": 0.9579539837683563 |
| }, |
| { |
| "cos_sim": 0.9563582420349122, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04364180862903595, |
| "epoch": 0.5580693815987934, |
| "grad_norm": 0.6657865643501282, |
| "learning_rate": 1.295649132186141e-05, |
| "loss": 0.18018112182617188, |
| "sft_loss": 0.0034485697746276856, |
| "step": 370, |
| "token_acc": 0.9584288103053199 |
| }, |
| { |
| "cos_sim": 0.9531916618347168, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04680837094783783, |
| "epoch": 0.5731523378582202, |
| "grad_norm": 0.5820720195770264, |
| "learning_rate": 1.2826408616450027e-05, |
| "loss": 0.1861934781074524, |
| "sft_loss": 0.003478127717971802, |
| "step": 380, |
| "token_acc": 0.9566955266955267 |
| }, |
| { |
| "cos_sim": 0.9560998916625977, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0439001590013504, |
| "epoch": 0.5882352941176471, |
| "grad_norm": 0.8649207949638367, |
| "learning_rate": 1.2693009562016084e-05, |
| "loss": 0.182230806350708, |
| "sft_loss": 0.0034997045993804933, |
| "step": 390, |
| "token_acc": 0.9569879802026868 |
| }, |
| { |
| "cos_sim": 0.9577288627624512, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.042271199822425845, |
| "epoch": 0.6033182503770739, |
| "grad_norm": 1.019384503364563, |
| "learning_rate": 1.2556377215992019e-05, |
| "loss": 0.17852908372879028, |
| "sft_loss": 0.0034654736518859863, |
| "step": 400, |
| "token_acc": 0.9581241117953576 |
| }, |
| { |
| "cos_sim": 0.9568939208984375, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043106073141098024, |
| "epoch": 0.6184012066365008, |
| "grad_norm": 0.8741854429244995, |
| "learning_rate": 1.2416596648934716e-05, |
| "loss": 0.17971467971801758, |
| "sft_loss": 0.0034607797861099243, |
| "step": 410, |
| "token_acc": 0.957463580274986 |
| }, |
| { |
| "cos_sim": 0.9562771797180176, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043722710013389586, |
| "epoch": 0.6334841628959276, |
| "grad_norm": 0.9732675552368164, |
| "learning_rate": 1.2273754891558542e-05, |
| "loss": 0.17926173210144042, |
| "sft_loss": 0.0034157931804656982, |
| "step": 420, |
| "token_acc": 0.9579150638323959 |
| }, |
| { |
| "cos_sim": 0.9562744140625, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043725624680519104, |
| "epoch": 0.6485671191553545, |
| "grad_norm": 0.8504221439361572, |
| "learning_rate": 1.212794088054797e-05, |
| "loss": 0.17900450229644777, |
| "sft_loss": 0.003407609462738037, |
| "step": 430, |
| "token_acc": 0.9575951205344176 |
| }, |
| { |
| "cos_sim": 0.9573652267456054, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04263473153114319, |
| "epoch": 0.6636500754147813, |
| "grad_norm": 0.6911851167678833, |
| "learning_rate": 1.1979245403183451e-05, |
| "loss": 0.1755327343940735, |
| "sft_loss": 0.003353661298751831, |
| "step": 440, |
| "token_acc": 0.9577667524194536 |
| }, |
| { |
| "cos_sim": 0.9589430809020996, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041057008504867556, |
| "epoch": 0.6787330316742082, |
| "grad_norm": 0.8256673216819763, |
| "learning_rate": 1.182776104081511e-05, |
| "loss": 0.17140830755233766, |
| "sft_loss": 0.0033036589622497557, |
| "step": 450, |
| "token_acc": 0.9586852351003294 |
| }, |
| { |
| "cos_sim": 0.9559874534606934, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04401255249977112, |
| "epoch": 0.693815987933635, |
| "grad_norm": 0.9640001654624939, |
| "learning_rate": 1.167358211121938e-05, |
| "loss": 0.17596397399902344, |
| "sft_loss": 0.0032982468605041505, |
| "step": 460, |
| "token_acc": 0.9592674651007855 |
| }, |
| { |
| "cos_sim": 0.9567716598510743, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043228423595428465, |
| "epoch": 0.7088989441930619, |
| "grad_norm": 0.8248547315597534, |
| "learning_rate": 1.1516804609874507e-05, |
| "loss": 0.17867534160614013, |
| "sft_loss": 0.00342218279838562, |
| "step": 470, |
| "token_acc": 0.9576837082276677 |
| }, |
| { |
| "cos_sim": 0.9568825721740722, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04311748445034027, |
| "epoch": 0.7239819004524887, |
| "grad_norm": 0.9829240441322327, |
| "learning_rate": 1.1357526150191471e-05, |
| "loss": 0.174727201461792, |
| "sft_loss": 0.0033043503761291506, |
| "step": 480, |
| "token_acc": 0.9591350273419852 |
| }, |
| { |
| "cos_sim": 0.9550409317016602, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04495901167392731, |
| "epoch": 0.7390648567119156, |
| "grad_norm": 0.7202755212783813, |
| "learning_rate": 1.1195845902737558e-05, |
| "loss": 0.1809255599975586, |
| "sft_loss": 0.0034059733152389526, |
| "step": 490, |
| "token_acc": 0.9573326718243693 |
| }, |
| { |
| "cos_sim": 0.9572129249572754, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0427870899438858, |
| "epoch": 0.7541478129713424, |
| "grad_norm": 0.667961835861206, |
| "learning_rate": 1.1031864533490397e-05, |
| "loss": 0.17608168125152587, |
| "sft_loss": 0.0033631980419158935, |
| "step": 500, |
| "token_acc": 0.9583889765691392 |
| }, |
| { |
| "cos_sim": 0.9579586029052735, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04204140901565552, |
| "epoch": 0.7692307692307693, |
| "grad_norm": 0.7028427124023438, |
| "learning_rate": 1.0865684141160912e-05, |
| "loss": 0.17287062406539916, |
| "sft_loss": 0.00330013632774353, |
| "step": 510, |
| "token_acc": 0.958580794446587 |
| }, |
| { |
| "cos_sim": 0.9577391624450684, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04226081669330597, |
| "epoch": 0.7843137254901961, |
| "grad_norm": 0.8001043796539307, |
| "learning_rate": 1.0697408193624248e-05, |
| "loss": 0.17270078659057617, |
| "sft_loss": 0.003283858299255371, |
| "step": 520, |
| "token_acc": 0.9596215576955567 |
| }, |
| { |
| "cos_sim": 0.9563358306884766, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04366412460803985, |
| "epoch": 0.799396681749623, |
| "grad_norm": 0.6969711780548096, |
| "learning_rate": 1.0527141463498186e-05, |
| "loss": 0.1776696562767029, |
| "sft_loss": 0.00336897075176239, |
| "step": 530, |
| "token_acc": 0.9582172289085196 |
| }, |
| { |
| "cos_sim": 0.9588777542114257, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041122305393219, |
| "epoch": 0.8144796380090498, |
| "grad_norm": 0.6015124320983887, |
| "learning_rate": 1.0354989962909223e-05, |
| "loss": 0.1682189702987671, |
| "sft_loss": 0.003200727701187134, |
| "step": 540, |
| "token_acc": 0.9603295649302662 |
| }, |
| { |
| "cos_sim": 0.9579531669616699, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0420468658208847, |
| "epoch": 0.8295625942684767, |
| "grad_norm": 0.6517713665962219, |
| "learning_rate": 1.0181060877486896e-05, |
| "loss": 0.17154401540756226, |
| "sft_loss": 0.0032584071159362793, |
| "step": 550, |
| "token_acc": 0.9595002540574469 |
| }, |
| { |
| "cos_sim": 0.9611032485961915, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03889682590961456, |
| "epoch": 0.8446455505279035, |
| "grad_norm": 0.791344165802002, |
| "learning_rate": 1.0005462499627443e-05, |
| "loss": 0.16251996755599976, |
| "sft_loss": 0.003133907914161682, |
| "step": 560, |
| "token_acc": 0.961120714961349 |
| }, |
| { |
| "cos_sim": 0.9566350936889648, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043365010619163515, |
| "epoch": 0.8597285067873304, |
| "grad_norm": 0.594643771648407, |
| "learning_rate": 9.82830416106839e-06, |
| "loss": 0.17398338317871093, |
| "sft_loss": 0.0032687306404113768, |
| "step": 570, |
| "token_acc": 0.9590278884462151 |
| }, |
| { |
| "cos_sim": 0.9601497650146484, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03985028862953186, |
| "epoch": 0.8748114630467572, |
| "grad_norm": 0.5415398478507996, |
| "learning_rate": 9.649696164815983e-06, |
| "loss": 0.16367310285568237, |
| "sft_loss": 0.003122270107269287, |
| "step": 580, |
| "token_acc": 0.9611103391845368 |
| }, |
| { |
| "cos_sim": 0.9599147796630859, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04008529782295227, |
| "epoch": 0.889894419306184, |
| "grad_norm": 0.6488097310066223, |
| "learning_rate": 9.469749716467922e-06, |
| "loss": 0.16612260341644286, |
| "sft_loss": 0.003187066316604614, |
| "step": 590, |
| "token_acc": 0.9604768853426061 |
| }, |
| { |
| "cos_sim": 0.9572768211364746, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04272326231002808, |
| "epoch": 0.9049773755656109, |
| "grad_norm": 0.6474986672401428, |
| "learning_rate": 9.288576854974107e-06, |
| "loss": 0.17227413654327392, |
| "sft_loss": 0.0032474040985107423, |
| "step": 600, |
| "token_acc": 0.9598335339239882 |
| }, |
| { |
| "cos_sim": 0.9604640960693359, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03953596353530884, |
| "epoch": 0.9200603318250377, |
| "grad_norm": 0.6288496851921082, |
| "learning_rate": 9.106290382878496e-06, |
| "loss": 0.16618789434432985, |
| "sft_loss": 0.0032165735960006715, |
| "step": 610, |
| "token_acc": 0.9597207796175231 |
| }, |
| { |
| "cos_sim": 0.9617090225219727, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.038291019201278684, |
| "epoch": 0.9351432880844646, |
| "grad_norm": 0.6721214652061462, |
| "learning_rate": 8.923003796085571e-06, |
| "loss": 0.16136680841445922, |
| "sft_loss": 0.0031281620264053344, |
| "step": 620, |
| "token_acc": 0.961438223938224 |
| }, |
| { |
| "cos_sim": 0.9553379058837891, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04466201663017273, |
| "epoch": 0.9502262443438914, |
| "grad_norm": 0.5502645969390869, |
| "learning_rate": 8.738831213195076e-06, |
| "loss": 0.17655622959136963, |
| "sft_loss": 0.0032842814922332763, |
| "step": 630, |
| "token_acc": 0.9586743281676243 |
| }, |
| { |
| "cos_sim": 0.958277702331543, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041722333431243895, |
| "epoch": 0.9653092006033183, |
| "grad_norm": 0.6368167996406555, |
| "learning_rate": 8.553887304449051e-06, |
| "loss": 0.16567234992980956, |
| "sft_loss": 0.0030911445617675783, |
| "step": 640, |
| "token_acc": 0.9614270749892169 |
| }, |
| { |
| "cos_sim": 0.9605836868286133, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.039416351914405824, |
| "epoch": 0.9803921568627451, |
| "grad_norm": 0.6817902326583862, |
| "learning_rate": 8.368287220335405e-06, |
| "loss": 0.16319282054901124, |
| "sft_loss": 0.0031289577484130858, |
| "step": 650, |
| "token_acc": 0.9607986668120299 |
| }, |
| { |
| "cos_sim": 0.9586556434631348, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.0413442999124527, |
| "epoch": 0.995475113122172, |
| "grad_norm": 0.7272047400474548, |
| "learning_rate": 8.18214651989248e-06, |
| "loss": 0.16760696172714235, |
| "sft_loss": 0.003170502185821533, |
| "step": 660, |
| "token_acc": 0.9606587660660332 |
| }, |
| { |
| "cos_sim": 0.961158561706543, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03884148597717285, |
| "epoch": 1.0105580693815988, |
| "grad_norm": 0.4904056191444397, |
| "learning_rate": 7.99558109875924e-06, |
| "loss": 0.1609166979789734, |
| "sft_loss": 0.0030865728855133056, |
| "step": 670, |
| "token_acc": 0.9618842421799919 |
| }, |
| { |
| "cos_sim": 0.9569988250732422, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04300126731395722, |
| "epoch": 1.0256410256410255, |
| "grad_norm": 0.7804086804389954, |
| "learning_rate": 7.808707117015871e-06, |
| "loss": 0.1645439624786377, |
| "sft_loss": 0.002991935610771179, |
| "step": 680, |
| "token_acc": 0.9634873880474275 |
| }, |
| { |
| "cos_sim": 0.9613571166992188, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03864296674728394, |
| "epoch": 1.0407239819004526, |
| "grad_norm": 0.9397616982460022, |
| "learning_rate": 7.621640926859767e-06, |
| "loss": 0.1597062587738037, |
| "sft_loss": 0.0030586719512939453, |
| "step": 690, |
| "token_acc": 0.9618396829854285 |
| }, |
| { |
| "cos_sim": 0.9568060874938965, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04319390058517456, |
| "epoch": 1.0558069381598794, |
| "grad_norm": 0.723336398601532, |
| "learning_rate": 7.434499000161847e-06, |
| "loss": 0.16262754201889038, |
| "sft_loss": 0.0029224157333374023, |
| "step": 700, |
| "token_acc": 0.9633835241413639 |
| }, |
| { |
| "cos_sim": 0.9574602127075196, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04253982305526734, |
| "epoch": 1.0708898944193062, |
| "grad_norm": 0.6005541086196899, |
| "learning_rate": 7.247397855948417e-06, |
| "loss": 0.1620495080947876, |
| "sft_loss": 0.0029370561242103577, |
| "step": 710, |
| "token_acc": 0.9636207473832851 |
| }, |
| { |
| "cos_sim": 0.9617099761962891, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03829004764556885, |
| "epoch": 1.085972850678733, |
| "grad_norm": 0.7606406211853027, |
| "learning_rate": 7.060453987853635e-06, |
| "loss": 0.15295138359069824, |
| "sft_loss": 0.002865228056907654, |
| "step": 720, |
| "token_acc": 0.9648746675263964 |
| }, |
| { |
| "cos_sim": 0.957455825805664, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04254419207572937, |
| "epoch": 1.10105580693816, |
| "grad_norm": 0.542752206325531, |
| "learning_rate": 6.873783791587794e-06, |
| "loss": 0.1618901252746582, |
| "sft_loss": 0.0029318571090698243, |
| "step": 730, |
| "token_acc": 0.9640453202570585 |
| }, |
| { |
| "cos_sim": 0.9582402229309082, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04175975024700165, |
| "epoch": 1.1161387631975868, |
| "grad_norm": 0.5904515981674194, |
| "learning_rate": 6.6875034924665825e-06, |
| "loss": 0.15857551097869874, |
| "sft_loss": 0.002867497503757477, |
| "step": 740, |
| "token_acc": 0.9638847578222961 |
| }, |
| { |
| "cos_sim": 0.9587013244628906, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041298767924308775, |
| "epoch": 1.1312217194570136, |
| "grad_norm": 0.5564361214637756, |
| "learning_rate": 6.501729073046421e-06, |
| "loss": 0.15704817771911622, |
| "sft_loss": 0.002842816710472107, |
| "step": 750, |
| "token_acc": 0.96492840783385 |
| }, |
| { |
| "cos_sim": 0.9581723213195801, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041827693581581116, |
| "epoch": 1.1463046757164403, |
| "grad_norm": 0.4957558214664459, |
| "learning_rate": 6.316576200910982e-06, |
| "loss": 0.15571348667144774, |
| "sft_loss": 0.0027746617794036864, |
| "step": 760, |
| "token_acc": 0.965580921450632 |
| }, |
| { |
| "cos_sim": 0.9609840393066407, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.039015993475914, |
| "epoch": 1.1613876319758674, |
| "grad_norm": 0.5658223032951355, |
| "learning_rate": 6.1321601566537614e-06, |
| "loss": 0.15808570384979248, |
| "sft_loss": 0.0029893785715103148, |
| "step": 770, |
| "token_acc": 0.963050224112171 |
| }, |
| { |
| "cos_sim": 0.9590757369995118, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.040924301743507384, |
| "epoch": 1.1764705882352942, |
| "grad_norm": 0.8099390268325806, |
| "learning_rate": 5.94859576210165e-06, |
| "loss": 0.1593359351158142, |
| "sft_loss": 0.002933032810688019, |
| "step": 780, |
| "token_acc": 0.9637189262909925 |
| }, |
| { |
| "cos_sim": 0.9608880043029785, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03911195397377014, |
| "epoch": 1.191553544494721, |
| "grad_norm": 0.594108521938324, |
| "learning_rate": 5.765997308824127e-06, |
| "loss": 0.15664927959442138, |
| "sft_loss": 0.0029396921396255492, |
| "step": 790, |
| "token_acc": 0.9627525942108137 |
| }, |
| { |
| "cos_sim": 0.962888240814209, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03711185157299042, |
| "epoch": 1.2066365007541477, |
| "grad_norm": 0.5782893300056458, |
| "learning_rate": 5.584478486972594e-06, |
| "loss": 0.1509794235229492, |
| "sft_loss": 0.0028625145554542542, |
| "step": 800, |
| "token_acc": 0.9644990612732548 |
| }, |
| { |
| "cos_sim": 0.960543441772461, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03945652544498444, |
| "epoch": 1.2217194570135748, |
| "grad_norm": 0.6433768272399902, |
| "learning_rate": 5.404152314494185e-06, |
| "loss": 0.15497575998306273, |
| "sft_loss": 0.002870166301727295, |
| "step": 810, |
| "token_acc": 0.9647309912787126 |
| }, |
| { |
| "cos_sim": 0.9622140884399414, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03778596818447113, |
| "epoch": 1.2368024132730016, |
| "grad_norm": 0.6467146873474121, |
| "learning_rate": 5.225131066764096e-06, |
| "loss": 0.15006879568099976, |
| "sft_loss": 0.00280035138130188, |
| "step": 820, |
| "token_acc": 0.9652448178082399 |
| }, |
| { |
| "cos_sim": 0.9583595275878907, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04164048433303833, |
| "epoch": 1.2518853695324283, |
| "grad_norm": 0.5411479473114014, |
| "learning_rate": 5.04752620668025e-06, |
| "loss": 0.1611725091934204, |
| "sft_loss": 0.002954617142677307, |
| "step": 830, |
| "token_acc": 0.9629328772979906 |
| }, |
| { |
| "cos_sim": 0.9565027236938477, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043497335910797116, |
| "epoch": 1.2669683257918551, |
| "grad_norm": 0.7859143614768982, |
| "learning_rate": 4.871448315263849e-06, |
| "loss": 0.15971229076385499, |
| "sft_loss": 0.0028161421418190003, |
| "step": 840, |
| "token_acc": 0.9648385764536664 |
| }, |
| { |
| "cos_sim": 0.9596900939941406, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04030995666980743, |
| "epoch": 1.282051282051282, |
| "grad_norm": 0.606879711151123, |
| "learning_rate": 4.697007022808989e-06, |
| "loss": 0.15423134565353394, |
| "sft_loss": 0.0028042316436767576, |
| "step": 850, |
| "token_acc": 0.9645077260079332 |
| }, |
| { |
| "cos_sim": 0.9622067451477051, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.037793251872062686, |
| "epoch": 1.297134238310709, |
| "grad_norm": 0.7979101538658142, |
| "learning_rate": 4.52431094062422e-06, |
| "loss": 0.15060091018676758, |
| "sft_loss": 0.0028166159987449645, |
| "step": 860, |
| "token_acc": 0.9646602300549729 |
| }, |
| { |
| "cos_sim": 0.9580653190612793, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041934633255004884, |
| "epoch": 1.3122171945701357, |
| "grad_norm": 0.6560086607933044, |
| "learning_rate": 4.353467593408566e-06, |
| "loss": 0.15815997123718262, |
| "sft_loss": 0.002845767140388489, |
| "step": 870, |
| "token_acc": 0.964312429871587 |
| }, |
| { |
| "cos_sim": 0.9604076385498047, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03959237039089203, |
| "epoch": 1.3273001508295625, |
| "grad_norm": 1.0168495178222656, |
| "learning_rate": 4.1845833523040775e-06, |
| "loss": 0.15515239238739015, |
| "sft_loss": 0.0028688937425613405, |
| "step": 880, |
| "token_acc": 0.9635724852071006 |
| }, |
| { |
| "cos_sim": 0.961815071105957, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03818497359752655, |
| "epoch": 1.3423831070889896, |
| "grad_norm": 0.5989195704460144, |
| "learning_rate": 4.0177633686666225e-06, |
| "loss": 0.14794611930847168, |
| "sft_loss": 0.0027140676975250244, |
| "step": 890, |
| "token_acc": 0.966539761725032 |
| }, |
| { |
| "cos_sim": 0.9618725776672363, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03812746405601501, |
| "epoch": 1.3574660633484164, |
| "grad_norm": 0.6104589700698853, |
| "learning_rate": 3.853111508596134e-06, |
| "loss": 0.14966075420379638, |
| "sft_loss": 0.0027705252170562742, |
| "step": 900, |
| "token_acc": 0.9650014664190047 |
| }, |
| { |
| "cos_sim": 0.9612789154052734, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03872108459472656, |
| "epoch": 1.3725490196078431, |
| "grad_norm": 0.667165219783783, |
| "learning_rate": 3.6907302882671214e-06, |
| "loss": 0.15173587799072266, |
| "sft_loss": 0.0028056919574737547, |
| "step": 910, |
| "token_acc": 0.9655432747478407 |
| }, |
| { |
| "cos_sim": 0.957188606262207, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.042811453342437744, |
| "epoch": 1.38763197586727, |
| "grad_norm": 0.6276723146438599, |
| "learning_rate": 3.5307208100996376e-06, |
| "loss": 0.1576075553894043, |
| "sft_loss": 0.002784663438796997, |
| "step": 920, |
| "token_acc": 0.9652169856794116 |
| }, |
| { |
| "cos_sim": 0.9586119651794434, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041388103365898134, |
| "epoch": 1.4027149321266967, |
| "grad_norm": 0.7700992822647095, |
| "learning_rate": 3.373182699810508e-06, |
| "loss": 0.15658562183380126, |
| "sft_loss": 0.002823895215988159, |
| "step": 930, |
| "token_acc": 0.9648340535074009 |
| }, |
| { |
| "cos_sim": 0.9586081504821777, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04139181971549988, |
| "epoch": 1.4177978883861238, |
| "grad_norm": 0.5908275246620178, |
| "learning_rate": 3.2182140443839834e-06, |
| "loss": 0.15772662162780762, |
| "sft_loss": 0.002859365940093994, |
| "step": 940, |
| "token_acc": 0.9637626109152789 |
| }, |
| { |
| "cos_sim": 0.956392478942871, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043607497215271, |
| "epoch": 1.4328808446455505, |
| "grad_norm": 0.6041402816772461, |
| "learning_rate": 3.0659113310004437e-06, |
| "loss": 0.16355013847351074, |
| "sft_loss": 0.002930566668510437, |
| "step": 950, |
| "token_acc": 0.9637660697035 |
| }, |
| { |
| "cos_sim": 0.9604990959167481, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03950095772743225, |
| "epoch": 1.4479638009049773, |
| "grad_norm": 0.5494619011878967, |
| "learning_rate": 2.9163693869611806e-06, |
| "loss": 0.15139245986938477, |
| "sft_loss": 0.002755966782569885, |
| "step": 960, |
| "token_acc": 0.9654504656810008 |
| }, |
| { |
| "cos_sim": 0.9605469703674316, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03945311903953552, |
| "epoch": 1.4630467571644044, |
| "grad_norm": 0.5176012516021729, |
| "learning_rate": 2.7696813206466494e-06, |
| "loss": 0.1496511936187744, |
| "sft_loss": 0.0027039438486099244, |
| "step": 970, |
| "token_acc": 0.967034284344282 |
| }, |
| { |
| "cos_sim": 0.9582103729248047, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04178963303565979, |
| "epoch": 1.4781297134238311, |
| "grad_norm": 0.5375176668167114, |
| "learning_rate": 2.6259384635449823e-06, |
| "loss": 0.15224893093109132, |
| "sft_loss": 0.002668297290802002, |
| "step": 980, |
| "token_acc": 0.966408741968469 |
| }, |
| { |
| "cos_sim": 0.9583855628967285, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.041614392399787904, |
| "epoch": 1.493212669683258, |
| "grad_norm": 0.6213409900665283, |
| "learning_rate": 2.48523031338683e-06, |
| "loss": 0.15863910913467408, |
| "sft_loss": 0.0028767526149749757, |
| "step": 990, |
| "token_acc": 0.964980719890771 |
| }, |
| { |
| "cos_sim": 0.9614453315734863, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03855462372303009, |
| "epoch": 1.5082956259426847, |
| "grad_norm": 0.6084774732589722, |
| "learning_rate": 2.347644478421916e-06, |
| "loss": 0.1520156145095825, |
| "sft_loss": 0.0028227567672729492, |
| "step": 1000, |
| "token_acc": 0.9643499144340423 |
| }, |
| { |
| "cos_sim": 0.9563823699951172, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.043617653846740725, |
| "epoch": 1.5233785822021115, |
| "grad_norm": 0.5677129030227661, |
| "learning_rate": 2.2132666228720783e-06, |
| "loss": 0.16164510250091552, |
| "sft_loss": 0.002870526909828186, |
| "step": 1010, |
| "token_acc": 0.9640724946695096 |
| }, |
| { |
| "cos_sim": 0.9657289505004882, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03427100479602814, |
| "epoch": 1.5384615384615383, |
| "grad_norm": 0.5602020621299744, |
| "learning_rate": 2.0821804135946685e-06, |
| "loss": 0.14286451339721679, |
| "sft_loss": 0.00275096595287323, |
| "step": 1020, |
| "token_acc": 0.9654056305981509 |
| }, |
| { |
| "cos_sim": 0.9586674690246582, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04133264124393463, |
| "epoch": 1.5535444947209653, |
| "grad_norm": 0.5280652642250061, |
| "learning_rate": 1.954467467989575e-06, |
| "loss": 0.15333621501922606, |
| "sft_loss": 0.0027251243591308594, |
| "step": 1030, |
| "token_acc": 0.9665621259639713 |
| }, |
| { |
| "cos_sim": 0.959110164642334, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04088979959487915, |
| "epoch": 1.5686274509803921, |
| "grad_norm": 0.5409430265426636, |
| "learning_rate": 1.8302073031822736e-06, |
| "loss": 0.15340681076049806, |
| "sft_loss": 0.0027494728565216064, |
| "step": 1040, |
| "token_acc": 0.9654812506617858 |
| }, |
| { |
| "cos_sim": 0.961603832244873, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03839621543884277, |
| "epoch": 1.5837104072398192, |
| "grad_norm": 0.5610837340354919, |
| "learning_rate": 1.7094772865145908e-06, |
| "loss": 0.14765048027038574, |
| "sft_loss": 0.00269426703453064, |
| "step": 1050, |
| "token_acc": 0.9670046289622475 |
| }, |
| { |
| "cos_sim": 0.9625849723815918, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03741500377655029, |
| "epoch": 1.598793363499246, |
| "grad_norm": 0.6360013484954834, |
| "learning_rate": 1.5923525873739312e-06, |
| "loss": 0.14554901123046876, |
| "sft_loss": 0.0026776567101478577, |
| "step": 1060, |
| "token_acc": 0.9659362770727379 |
| }, |
| { |
| "cos_sim": 0.9599601745605468, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.040039819478988645, |
| "epoch": 1.6138763197586727, |
| "grad_norm": 0.6609951257705688, |
| "learning_rate": 1.4789061303910304e-06, |
| "loss": 0.15372539758682252, |
| "sft_loss": 0.002801927924156189, |
| "step": 1070, |
| "token_acc": 0.9649936158569952 |
| }, |
| { |
| "cos_sim": 0.9597688674926758, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04023122191429138, |
| "epoch": 1.6289592760180995, |
| "grad_norm": 0.6310714483261108, |
| "learning_rate": 1.3692085500353444e-06, |
| "loss": 0.1543022871017456, |
| "sft_loss": 0.0028103858232498167, |
| "step": 1080, |
| "token_acc": 0.9652636474623568 |
| }, |
| { |
| "cos_sim": 0.9623384475708008, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.037661558389663695, |
| "epoch": 1.6440422322775263, |
| "grad_norm": 0.5179605484008789, |
| "learning_rate": 1.263328146636342e-06, |
| "loss": 0.1467289924621582, |
| "sft_loss": 0.0027022033929824827, |
| "step": 1090, |
| "token_acc": 0.9658165947211648 |
| }, |
| { |
| "cos_sim": 0.9614273071289062, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03857271373271942, |
| "epoch": 1.6591251885369531, |
| "grad_norm": 0.47959592938423157, |
| "learning_rate": 1.1613308438580924e-06, |
| "loss": 0.149735426902771, |
| "sft_loss": 0.0027505964040756225, |
| "step": 1100, |
| "token_acc": 0.9648848166735731 |
| }, |
| { |
| "cos_sim": 0.9636540412902832, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03634599447250366, |
| "epoch": 1.6742081447963801, |
| "grad_norm": 0.527296781539917, |
| "learning_rate": 1.063280147653618e-06, |
| "loss": 0.14220519065856935, |
| "sft_loss": 0.0026266127824783325, |
| "step": 1110, |
| "token_acc": 0.9672569546358804 |
| }, |
| { |
| "cos_sim": 0.9571693420410157, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04283071160316467, |
| "epoch": 1.689291101055807, |
| "grad_norm": 0.4759458303451538, |
| "learning_rate": 9.692371067245726e-07, |
| "loss": 0.15967360734939576, |
| "sft_loss": 0.0028482645750045775, |
| "step": 1120, |
| "token_acc": 0.9646659011203873 |
| }, |
| { |
| "cos_sim": 0.9603818893432617, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03961814045906067, |
| "epoch": 1.704374057315234, |
| "grad_norm": 0.4827066659927368, |
| "learning_rate": 8.792602745108685e-07, |
| "loss": 0.14749515056610107, |
| "sft_loss": 0.002628315985202789, |
| "step": 1130, |
| "token_acc": 0.9664777353291163 |
| }, |
| { |
| "cos_sim": 0.9596490859985352, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.040350842475891116, |
| "epoch": 1.7194570135746607, |
| "grad_norm": 0.5767756700515747, |
| "learning_rate": 7.93405672733902e-07, |
| "loss": 0.15450209379196167, |
| "sft_loss": 0.0028106480836868285, |
| "step": 1140, |
| "token_acc": 0.9643665243125796 |
| }, |
| { |
| "cos_sim": 0.9617944717407226, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03820553421974182, |
| "epoch": 1.7345399698340875, |
| "grad_norm": 0.5325683951377869, |
| "learning_rate": 7.117267565160989e-07, |
| "loss": 0.14688180685043334, |
| "sft_loss": 0.002679780125617981, |
| "step": 1150, |
| "token_acc": 0.9659744184959571 |
| }, |
| { |
| "cos_sim": 0.9576915740966797, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04230844974517822, |
| "epoch": 1.7496229260935143, |
| "grad_norm": 0.5952834486961365, |
| "learning_rate": 6.342743810984836e-07, |
| "loss": 0.1566309928894043, |
| "sft_loss": 0.002779296040534973, |
| "step": 1160, |
| "token_acc": 0.9652321327662612 |
| }, |
| { |
| "cos_sim": 0.9652015686035156, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03479845523834228, |
| "epoch": 1.7647058823529411, |
| "grad_norm": 0.563335120677948, |
| "learning_rate": 5.610967701769814e-07, |
| "loss": 0.141930890083313, |
| "sft_loss": 0.0026954174041748046, |
| "step": 1170, |
| "token_acc": 0.9666916752633717 |
| }, |
| { |
| "cos_sim": 0.9610443115234375, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03895579874515533, |
| "epoch": 1.779788838612368, |
| "grad_norm": 0.776590883731842, |
| "learning_rate": 4.922394858772067e-07, |
| "loss": 0.15091207027435302, |
| "sft_loss": 0.0027682125568389893, |
| "step": 1180, |
| "token_acc": 0.9653806080889086 |
| }, |
| { |
| "cos_sim": 0.9644537925720215, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.035546216368675235, |
| "epoch": 1.7948717948717947, |
| "grad_norm": 0.482070654630661, |
| "learning_rate": 4.2774540038639286e-07, |
| "loss": 0.1390576958656311, |
| "sft_loss": 0.0025682419538497923, |
| "step": 1190, |
| "token_acc": 0.967479937349631 |
| }, |
| { |
| "cos_sim": 0.9611939430236817, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03880613148212433, |
| "epoch": 1.8099547511312217, |
| "grad_norm": 0.5717806816101074, |
| "learning_rate": 3.676546692601465e-07, |
| "loss": 0.14678553342819214, |
| "sft_loss": 0.0026467412710189818, |
| "step": 1200, |
| "token_acc": 0.9670608902970794 |
| }, |
| { |
| "cos_sim": 0.9606647491455078, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.039335194230079654, |
| "epoch": 1.8250377073906485, |
| "grad_norm": 0.4626607298851013, |
| "learning_rate": 3.1200470642063716e-07, |
| "loss": 0.14852668046951295, |
| "sft_loss": 0.002674698829650879, |
| "step": 1210, |
| "token_acc": 0.9666944341601876 |
| }, |
| { |
| "cos_sim": 0.9623315811157227, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03766842782497406, |
| "epoch": 1.8401206636500755, |
| "grad_norm": 0.5528730154037476, |
| "learning_rate": 2.6083016086180715e-07, |
| "loss": 0.14226431846618653, |
| "sft_loss": 0.00256233811378479, |
| "step": 1220, |
| "token_acc": 0.9690826869327639 |
| }, |
| { |
| "cos_sim": 0.9570770263671875, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04292306005954742, |
| "epoch": 1.8552036199095023, |
| "grad_norm": 0.4962717592716217, |
| "learning_rate": 2.14162895076071e-07, |
| "loss": 0.1580639362335205, |
| "sft_loss": 0.002793344855308533, |
| "step": 1230, |
| "token_acc": 0.9648224657206156 |
| }, |
| { |
| "cos_sim": 0.9596912384033203, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.040308722853660585, |
| "epoch": 1.8702865761689291, |
| "grad_norm": 0.5444379448890686, |
| "learning_rate": 1.7203196521597354e-07, |
| "loss": 0.14951192140579223, |
| "sft_loss": 0.0026568114757537843, |
| "step": 1240, |
| "token_acc": 0.9667831067472307 |
| }, |
| { |
| "cos_sim": 0.9610977172851562, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03890231251716614, |
| "epoch": 1.885369532428356, |
| "grad_norm": 0.4733203947544098, |
| "learning_rate": 1.3446360300313375e-07, |
| "loss": 0.14659295082092286, |
| "sft_loss": 0.0026359140872955324, |
| "step": 1250, |
| "token_acc": 0.9674186654001424 |
| }, |
| { |
| "cos_sim": 0.9640268325805664, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03597318828105926, |
| "epoch": 1.9004524886877827, |
| "grad_norm": 0.5802148580551147, |
| "learning_rate": 1.0148119939575207e-07, |
| "loss": 0.14078178405761718, |
| "sft_loss": 0.0026007711887359617, |
| "step": 1260, |
| "token_acc": 0.9674666666666667 |
| }, |
| { |
| "cos_sim": 0.9594644546508789, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04053559899330139, |
| "epoch": 1.9155354449472095, |
| "grad_norm": 0.4988519847393036, |
| "learning_rate": 7.310529002484079e-08, |
| "loss": 0.15229123830795288, |
| "sft_loss": 0.002732321619987488, |
| "step": 1270, |
| "token_acc": 0.9661774555196413 |
| }, |
| { |
| "cos_sim": 0.960145378112793, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03985467553138733, |
| "epoch": 1.9306184012066365, |
| "grad_norm": 0.5619922280311584, |
| "learning_rate": 4.9353542408249465e-08, |
| "loss": 0.15132608413696289, |
| "sft_loss": 0.002736206352710724, |
| "step": 1280, |
| "token_acc": 0.965960607138989 |
| }, |
| { |
| "cos_sim": 0.9620681762695312, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03793186545372009, |
| "epoch": 1.9457013574660633, |
| "grad_norm": 0.5861108303070068, |
| "learning_rate": 3.024074495044826e-08, |
| "loss": 0.1475370168685913, |
| "sft_loss": 0.0027139380574226378, |
| "step": 1290, |
| "token_acc": 0.966789667896679 |
| }, |
| { |
| "cos_sim": 0.9627554893493653, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.03724450767040253, |
| "epoch": 1.9607843137254903, |
| "grad_norm": 0.5047047734260559, |
| "learning_rate": 1.5778797734913307e-08, |
| "loss": 0.1461432695388794, |
| "sft_loss": 0.0027047514915466307, |
| "step": 1300, |
| "token_acc": 0.9653572346246277 |
| }, |
| { |
| "cos_sim": 0.9619269371032715, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.038073021173477176, |
| "epoch": 1.9758672699849171, |
| "grad_norm": 0.5808618664741516, |
| "learning_rate": 5.976705114852654e-09, |
| "loss": 0.14662744998931884, |
| "sft_loss": 0.0026784569025039675, |
| "step": 1310, |
| "token_acc": 0.9660818900662675 |
| }, |
| { |
| "cos_sim": 0.9581977844238281, |
| "distill_lambda": 0.05, |
| "distill_loss": 0.04180223047733307, |
| "epoch": 1.990950226244344, |
| "grad_norm": 0.513159990310669, |
| "learning_rate": 8.40570106880878e-10, |
| "loss": 0.15337657928466797, |
| "sft_loss": 0.0027029067277908325, |
| "step": 1320, |
| "token_acc": 0.9666289507430037 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1326, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.6265711502005109e+19, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
