yalhessi's picture
Training in progress, epoch 5, checkpoint
da04dad verified
raw
history blame
32.9 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 3118,
"global_step": 77935,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.032079042761364,
"grad_norm": 1.7145991325378418,
"learning_rate": 0.0007978742969568658,
"loss": 0.4787,
"step": 500
},
{
"epoch": 0.064158085522728,
"grad_norm": 1.2714065313339233,
"learning_rate": 0.0007957357627082398,
"loss": 0.3972,
"step": 1000
},
{
"epoch": 0.096237128284092,
"grad_norm": 1.1843117475509644,
"learning_rate": 0.0007935972284596139,
"loss": 0.3736,
"step": 1500
},
{
"epoch": 0.128316171045456,
"grad_norm": 1.1524124145507812,
"learning_rate": 0.0007914629712794851,
"loss": 0.3626,
"step": 2000
},
{
"epoch": 0.16039521380682,
"grad_norm": 1.0530339479446411,
"learning_rate": 0.0007893244370308591,
"loss": 0.355,
"step": 2500
},
{
"epoch": 0.192474256568184,
"grad_norm": 1.0044286251068115,
"learning_rate": 0.0007871859027822331,
"loss": 0.3527,
"step": 3000
},
{
"epoch": 0.2000449106598659,
"eval_loss": 0.3370795249938965,
"eval_runtime": 6.2984,
"eval_samples_per_second": 79.386,
"eval_steps_per_second": 10.003,
"step": 3118
},
{
"epoch": 0.224553299329548,
"grad_norm": 1.4645978212356567,
"learning_rate": 0.0007850473685336071,
"loss": 0.3394,
"step": 3500
},
{
"epoch": 0.256632342090912,
"grad_norm": 1.2185771465301514,
"learning_rate": 0.0007829088342849811,
"loss": 0.3398,
"step": 4000
},
{
"epoch": 0.288711384852276,
"grad_norm": 1.3087233304977417,
"learning_rate": 0.0007807703000363551,
"loss": 0.3399,
"step": 4500
},
{
"epoch": 0.32079042761364,
"grad_norm": 1.411337971687317,
"learning_rate": 0.0007786360428562264,
"loss": 0.3355,
"step": 5000
},
{
"epoch": 0.352869470375004,
"grad_norm": 1.1822813749313354,
"learning_rate": 0.0007764975086076004,
"loss": 0.3329,
"step": 5500
},
{
"epoch": 0.384948513136368,
"grad_norm": 0.9898410439491272,
"learning_rate": 0.0007743632514274716,
"loss": 0.3311,
"step": 6000
},
{
"epoch": 0.4000898213197318,
"eval_loss": 0.31450170278549194,
"eval_runtime": 9.8807,
"eval_samples_per_second": 50.604,
"eval_steps_per_second": 6.376,
"step": 6236
},
{
"epoch": 0.417027555897732,
"grad_norm": 0.9827210307121277,
"learning_rate": 0.0007722247171788456,
"loss": 0.3251,
"step": 6500
},
{
"epoch": 0.449106598659096,
"grad_norm": 1.2011586427688599,
"learning_rate": 0.0007700861829302197,
"loss": 0.3271,
"step": 7000
},
{
"epoch": 0.48118564142046,
"grad_norm": 1.3871015310287476,
"learning_rate": 0.0007679476486815936,
"loss": 0.3259,
"step": 7500
},
{
"epoch": 0.513264684181824,
"grad_norm": 1.1245856285095215,
"learning_rate": 0.0007658091144329678,
"loss": 0.3192,
"step": 8000
},
{
"epoch": 0.545343726943188,
"grad_norm": 1.0883513689041138,
"learning_rate": 0.0007636705801843417,
"loss": 0.3209,
"step": 8500
},
{
"epoch": 0.577422769704552,
"grad_norm": 1.0555607080459595,
"learning_rate": 0.0007615320459357157,
"loss": 0.3178,
"step": 9000
},
{
"epoch": 0.6001347319795978,
"eval_loss": 0.3085324764251709,
"eval_runtime": 7.9945,
"eval_samples_per_second": 62.543,
"eval_steps_per_second": 7.88,
"step": 9354
},
{
"epoch": 0.6095018124659161,
"grad_norm": 1.0429250001907349,
"learning_rate": 0.0007593935116870898,
"loss": 0.3169,
"step": 9500
},
{
"epoch": 0.64158085522728,
"grad_norm": 1.2145720720291138,
"learning_rate": 0.0007572549774384637,
"loss": 0.3152,
"step": 10000
},
{
"epoch": 0.673659897988644,
"grad_norm": 1.3464765548706055,
"learning_rate": 0.0007551207202583349,
"loss": 0.3125,
"step": 10500
},
{
"epoch": 0.705738940750008,
"grad_norm": 1.1744924783706665,
"learning_rate": 0.000752982186009709,
"loss": 0.3086,
"step": 11000
},
{
"epoch": 0.7378179835113721,
"grad_norm": 1.234157919883728,
"learning_rate": 0.000750843651761083,
"loss": 0.3133,
"step": 11500
},
{
"epoch": 0.769897026272736,
"grad_norm": 0.9248010516166687,
"learning_rate": 0.0007487051175124569,
"loss": 0.3122,
"step": 12000
},
{
"epoch": 0.8001796426394636,
"eval_loss": 0.3022182583808899,
"eval_runtime": 8.0646,
"eval_samples_per_second": 61.999,
"eval_steps_per_second": 7.812,
"step": 12472
},
{
"epoch": 0.8019760690341,
"grad_norm": 0.862856388092041,
"learning_rate": 0.000746566583263831,
"loss": 0.308,
"step": 12500
},
{
"epoch": 0.834055111795464,
"grad_norm": 1.0224480628967285,
"learning_rate": 0.000744428049015205,
"loss": 0.31,
"step": 13000
},
{
"epoch": 0.8661341545568281,
"grad_norm": 1.008195161819458,
"learning_rate": 0.0007422937918350763,
"loss": 0.3079,
"step": 13500
},
{
"epoch": 0.898213197318192,
"grad_norm": 1.29293954372406,
"learning_rate": 0.0007401552575864503,
"loss": 0.3047,
"step": 14000
},
{
"epoch": 0.930292240079556,
"grad_norm": 0.9913870692253113,
"learning_rate": 0.0007380167233378243,
"loss": 0.3074,
"step": 14500
},
{
"epoch": 0.96237128284092,
"grad_norm": 1.1802239418029785,
"learning_rate": 0.0007358781890891983,
"loss": 0.3055,
"step": 15000
},
{
"epoch": 0.9944503256022841,
"grad_norm": 1.0421093702316284,
"learning_rate": 0.0007337396548405722,
"loss": 0.3106,
"step": 15500
},
{
"epoch": 1.0001924742565682,
"eval_loss": 0.30451422929763794,
"eval_runtime": 6.8665,
"eval_samples_per_second": 72.817,
"eval_steps_per_second": 9.175,
"step": 15590
},
{
"epoch": 1.0264972893208866,
"grad_norm": 1.1343709230422974,
"learning_rate": 0.0007316011205919463,
"loss": 0.3028,
"step": 16000
},
{
"epoch": 1.0585763320822508,
"grad_norm": 0.992735743522644,
"learning_rate": 0.0007294625863433203,
"loss": 0.3026,
"step": 16500
},
{
"epoch": 1.0906553748436147,
"grad_norm": 0.9756571650505066,
"learning_rate": 0.0007273240520946943,
"loss": 0.3004,
"step": 17000
},
{
"epoch": 1.1227344176049787,
"grad_norm": 1.005118489265442,
"learning_rate": 0.0007251897949145656,
"loss": 0.2953,
"step": 17500
},
{
"epoch": 1.1548134603663427,
"grad_norm": 1.2004996538162231,
"learning_rate": 0.0007230512606659396,
"loss": 0.3039,
"step": 18000
},
{
"epoch": 1.1868925031277067,
"grad_norm": 1.2973647117614746,
"learning_rate": 0.0007209127264173135,
"loss": 0.2997,
"step": 18500
},
{
"epoch": 1.200237384916434,
"eval_loss": 0.2830166518688202,
"eval_runtime": 6.4164,
"eval_samples_per_second": 77.926,
"eval_steps_per_second": 9.819,
"step": 18708
},
{
"epoch": 1.2189715458890706,
"grad_norm": 1.2329021692276,
"learning_rate": 0.0007187741921686877,
"loss": 0.2972,
"step": 19000
},
{
"epoch": 1.2510505886504346,
"grad_norm": 0.8207685947418213,
"learning_rate": 0.0007166399349885588,
"loss": 0.2979,
"step": 19500
},
{
"epoch": 1.2831296314117986,
"grad_norm": 1.1549195051193237,
"learning_rate": 0.000714501400739933,
"loss": 0.2982,
"step": 20000
},
{
"epoch": 1.3152086741731628,
"grad_norm": 1.0019187927246094,
"learning_rate": 0.0007123628664913069,
"loss": 0.2975,
"step": 20500
},
{
"epoch": 1.3472877169345265,
"grad_norm": 1.023240327835083,
"learning_rate": 0.0007102243322426809,
"loss": 0.2957,
"step": 21000
},
{
"epoch": 1.3793667596958907,
"grad_norm": 1.0395528078079224,
"learning_rate": 0.0007080900750625522,
"loss": 0.2957,
"step": 21500
},
{
"epoch": 1.4002822955763001,
"eval_loss": 0.28411924839019775,
"eval_runtime": 8.0825,
"eval_samples_per_second": 61.862,
"eval_steps_per_second": 7.795,
"step": 21826
},
{
"epoch": 1.4114458024572547,
"grad_norm": 1.2371494770050049,
"learning_rate": 0.0007059515408139262,
"loss": 0.2948,
"step": 22000
},
{
"epoch": 1.4435248452186187,
"grad_norm": 1.1509361267089844,
"learning_rate": 0.0007038130065653001,
"loss": 0.2939,
"step": 22500
},
{
"epoch": 1.4756038879799827,
"grad_norm": 0.9148961305618286,
"learning_rate": 0.0007016744723166742,
"loss": 0.2945,
"step": 23000
},
{
"epoch": 1.5076829307413466,
"grad_norm": 1.301382064819336,
"learning_rate": 0.0006995359380680482,
"loss": 0.2935,
"step": 23500
},
{
"epoch": 1.5397619735027108,
"grad_norm": 1.1036344766616821,
"learning_rate": 0.0006974059579564167,
"loss": 0.2946,
"step": 24000
},
{
"epoch": 1.5718410162640746,
"grad_norm": 1.6663960218429565,
"learning_rate": 0.0006952674237077907,
"loss": 0.2922,
"step": 24500
},
{
"epoch": 1.600327206236166,
"eval_loss": 0.2768917977809906,
"eval_runtime": 6.447,
"eval_samples_per_second": 77.556,
"eval_steps_per_second": 9.772,
"step": 24944
},
{
"epoch": 1.6039200590254388,
"grad_norm": 1.0343741178512573,
"learning_rate": 0.0006931288894591647,
"loss": 0.2928,
"step": 25000
},
{
"epoch": 1.6359991017868025,
"grad_norm": 1.3965007066726685,
"learning_rate": 0.0006909903552105388,
"loss": 0.2923,
"step": 25500
},
{
"epoch": 1.6680781445481667,
"grad_norm": 1.1331425905227661,
"learning_rate": 0.0006888518209619127,
"loss": 0.2936,
"step": 26000
},
{
"epoch": 1.7001571873095307,
"grad_norm": 1.0256651639938354,
"learning_rate": 0.0006867132867132868,
"loss": 0.2901,
"step": 26500
},
{
"epoch": 1.7322362300708947,
"grad_norm": 1.1987167596817017,
"learning_rate": 0.0006845747524646608,
"loss": 0.2932,
"step": 27000
},
{
"epoch": 1.7643152728322586,
"grad_norm": 0.9224876165390015,
"learning_rate": 0.0006824362182160348,
"loss": 0.2957,
"step": 27500
},
{
"epoch": 1.7963943155936226,
"grad_norm": 1.1595081090927124,
"learning_rate": 0.0006802976839674088,
"loss": 0.2895,
"step": 28000
},
{
"epoch": 1.8003721168960318,
"eval_loss": 0.2766253650188446,
"eval_runtime": 6.995,
"eval_samples_per_second": 71.48,
"eval_steps_per_second": 9.006,
"step": 28062
},
{
"epoch": 1.8284733583549868,
"grad_norm": 1.0937904119491577,
"learning_rate": 0.0006781634267872801,
"loss": 0.2908,
"step": 28500
},
{
"epoch": 1.8605524011163506,
"grad_norm": 1.6129448413848877,
"learning_rate": 0.0006760291696071513,
"loss": 0.2865,
"step": 29000
},
{
"epoch": 1.8926314438777148,
"grad_norm": 1.5731992721557617,
"learning_rate": 0.0006738906353585254,
"loss": 0.2865,
"step": 29500
},
{
"epoch": 1.9247104866390787,
"grad_norm": 1.1734341382980347,
"learning_rate": 0.0006717521011098993,
"loss": 0.2858,
"step": 30000
},
{
"epoch": 1.9567895294004427,
"grad_norm": 1.277829647064209,
"learning_rate": 0.0006696135668612733,
"loss": 0.2878,
"step": 30500
},
{
"epoch": 1.9888685721618067,
"grad_norm": 1.2518751621246338,
"learning_rate": 0.0006674793096811446,
"loss": 0.2847,
"step": 31000
},
{
"epoch": 2.0003849485131364,
"eval_loss": 0.2660098075866699,
"eval_runtime": 8.1469,
"eval_samples_per_second": 61.373,
"eval_steps_per_second": 7.733,
"step": 31180
},
{
"epoch": 2.0209155358804094,
"grad_norm": 1.0406543016433716,
"learning_rate": 0.0006653407754325186,
"loss": 0.2821,
"step": 31500
},
{
"epoch": 2.052994578641773,
"grad_norm": 1.1324297189712524,
"learning_rate": 0.0006632022411838926,
"loss": 0.2848,
"step": 32000
},
{
"epoch": 2.0850736214031373,
"grad_norm": 1.0970991849899292,
"learning_rate": 0.0006610637069352665,
"loss": 0.2815,
"step": 32500
},
{
"epoch": 2.1171526641645015,
"grad_norm": 1.0462357997894287,
"learning_rate": 0.0006589251726866407,
"loss": 0.2825,
"step": 33000
},
{
"epoch": 2.1492317069258653,
"grad_norm": 1.122344732284546,
"learning_rate": 0.0006567866384380146,
"loss": 0.2816,
"step": 33500
},
{
"epoch": 2.1813107496872295,
"grad_norm": 1.108782410621643,
"learning_rate": 0.0006546481041893886,
"loss": 0.2826,
"step": 34000
},
{
"epoch": 2.2004298591730023,
"eval_loss": 0.2733038365840912,
"eval_runtime": 7.489,
"eval_samples_per_second": 66.765,
"eval_steps_per_second": 8.412,
"step": 34298
},
{
"epoch": 2.2133897924485932,
"grad_norm": 1.092699646949768,
"learning_rate": 0.0006525095699407627,
"loss": 0.2836,
"step": 34500
},
{
"epoch": 2.2454688352099574,
"grad_norm": 1.0987275838851929,
"learning_rate": 0.0006503753127606339,
"loss": 0.2779,
"step": 35000
},
{
"epoch": 2.277547877971321,
"grad_norm": 1.1260478496551514,
"learning_rate": 0.0006482367785120078,
"loss": 0.2788,
"step": 35500
},
{
"epoch": 2.3096269207326854,
"grad_norm": 1.2035844326019287,
"learning_rate": 0.0006460982442633819,
"loss": 0.284,
"step": 36000
},
{
"epoch": 2.341705963494049,
"grad_norm": 1.0569308996200562,
"learning_rate": 0.0006439597100147559,
"loss": 0.2812,
"step": 36500
},
{
"epoch": 2.3737850062554133,
"grad_norm": 0.8856455683708191,
"learning_rate": 0.0006418254528346272,
"loss": 0.2767,
"step": 37000
},
{
"epoch": 2.400474769832868,
"eval_loss": 0.2658798098564148,
"eval_runtime": 8.0192,
"eval_samples_per_second": 62.35,
"eval_steps_per_second": 7.856,
"step": 37416
},
{
"epoch": 2.4058640490167775,
"grad_norm": 0.9563117027282715,
"learning_rate": 0.0006396911956544984,
"loss": 0.2787,
"step": 37500
},
{
"epoch": 2.4379430917781413,
"grad_norm": 0.9825711250305176,
"learning_rate": 0.0006375526614058725,
"loss": 0.2775,
"step": 38000
},
{
"epoch": 2.4700221345395055,
"grad_norm": 1.1553630828857422,
"learning_rate": 0.0006354141271572465,
"loss": 0.2779,
"step": 38500
},
{
"epoch": 2.5021011773008692,
"grad_norm": 1.1160862445831299,
"learning_rate": 0.0006332755929086204,
"loss": 0.2819,
"step": 39000
},
{
"epoch": 2.5341802200622334,
"grad_norm": 1.2663159370422363,
"learning_rate": 0.0006311370586599945,
"loss": 0.2797,
"step": 39500
},
{
"epoch": 2.566259262823597,
"grad_norm": 1.028053641319275,
"learning_rate": 0.0006289985244113685,
"loss": 0.2776,
"step": 40000
},
{
"epoch": 2.5983383055849614,
"grad_norm": 1.0482547283172607,
"learning_rate": 0.0006268642672312397,
"loss": 0.2737,
"step": 40500
},
{
"epoch": 2.600519680492734,
"eval_loss": 0.26375094056129456,
"eval_runtime": 7.2264,
"eval_samples_per_second": 69.19,
"eval_steps_per_second": 8.718,
"step": 40534
},
{
"epoch": 2.6304173483463256,
"grad_norm": 1.0435316562652588,
"learning_rate": 0.0006247257329826138,
"loss": 0.2783,
"step": 41000
},
{
"epoch": 2.6624963911076893,
"grad_norm": 0.866886556148529,
"learning_rate": 0.0006225871987339878,
"loss": 0.2738,
"step": 41500
},
{
"epoch": 2.694575433869053,
"grad_norm": 1.069162368774414,
"learning_rate": 0.0006204486644853617,
"loss": 0.2783,
"step": 42000
},
{
"epoch": 2.7266544766304173,
"grad_norm": 1.450871229171753,
"learning_rate": 0.0006183101302367357,
"loss": 0.275,
"step": 42500
},
{
"epoch": 2.7587335193917815,
"grad_norm": 1.0573506355285645,
"learning_rate": 0.000616175873056607,
"loss": 0.2762,
"step": 43000
},
{
"epoch": 2.790812562153145,
"grad_norm": 1.1295759677886963,
"learning_rate": 0.000614037338807981,
"loss": 0.2734,
"step": 43500
},
{
"epoch": 2.8005645911526003,
"eval_loss": 0.26422494649887085,
"eval_runtime": 8.0132,
"eval_samples_per_second": 62.397,
"eval_steps_per_second": 7.862,
"step": 43652
},
{
"epoch": 2.8228916049145094,
"grad_norm": 1.1608061790466309,
"learning_rate": 0.000611898804559355,
"loss": 0.272,
"step": 44000
},
{
"epoch": 2.8549706476758736,
"grad_norm": 1.2949804067611694,
"learning_rate": 0.000609760270310729,
"loss": 0.272,
"step": 44500
},
{
"epoch": 2.8870496904372374,
"grad_norm": 0.9073989391326904,
"learning_rate": 0.0006076217360621031,
"loss": 0.2744,
"step": 45000
},
{
"epoch": 2.919128733198601,
"grad_norm": 0.8703099489212036,
"learning_rate": 0.0006054874788819743,
"loss": 0.2762,
"step": 45500
},
{
"epoch": 2.9512077759599653,
"grad_norm": 1.0385627746582031,
"learning_rate": 0.0006033489446333484,
"loss": 0.2732,
"step": 46000
},
{
"epoch": 2.9832868187213295,
"grad_norm": 1.1464110612869263,
"learning_rate": 0.0006012104103847223,
"loss": 0.2701,
"step": 46500
},
{
"epoch": 3.0005774227697044,
"eval_loss": 0.2622199058532715,
"eval_runtime": 6.3897,
"eval_samples_per_second": 78.251,
"eval_steps_per_second": 9.86,
"step": 46770
},
{
"epoch": 3.015333782439932,
"grad_norm": 1.2512695789337158,
"learning_rate": 0.0005990718761360964,
"loss": 0.2697,
"step": 47000
},
{
"epoch": 3.047412825201296,
"grad_norm": 0.7882702946662903,
"learning_rate": 0.0005969376189559676,
"loss": 0.2685,
"step": 47500
},
{
"epoch": 3.07949186796266,
"grad_norm": 1.282387614250183,
"learning_rate": 0.0005947990847073416,
"loss": 0.2674,
"step": 48000
},
{
"epoch": 3.111570910724024,
"grad_norm": 1.4220795631408691,
"learning_rate": 0.0005926605504587157,
"loss": 0.2676,
"step": 48500
},
{
"epoch": 3.143649953485388,
"grad_norm": 1.0576750040054321,
"learning_rate": 0.0005905220162100896,
"loss": 0.269,
"step": 49000
},
{
"epoch": 3.175728996246752,
"grad_norm": 1.0533617734909058,
"learning_rate": 0.0005883834819614636,
"loss": 0.2629,
"step": 49500
},
{
"epoch": 3.2006223334295703,
"eval_loss": 0.2548312246799469,
"eval_runtime": 6.5962,
"eval_samples_per_second": 75.802,
"eval_steps_per_second": 9.551,
"step": 49888
},
{
"epoch": 3.207808039008116,
"grad_norm": 1.1989775896072388,
"learning_rate": 0.0005862449477128377,
"loss": 0.2666,
"step": 50000
},
{
"epoch": 3.23988708176948,
"grad_norm": 1.0356203317642212,
"learning_rate": 0.0005841064134642117,
"loss": 0.2622,
"step": 50500
},
{
"epoch": 3.2719661245308442,
"grad_norm": 1.0133503675460815,
"learning_rate": 0.0005819721562840828,
"loss": 0.2671,
"step": 51000
},
{
"epoch": 3.304045167292208,
"grad_norm": 1.1607108116149902,
"learning_rate": 0.000579833622035457,
"loss": 0.2633,
"step": 51500
},
{
"epoch": 3.336124210053572,
"grad_norm": 1.0809621810913086,
"learning_rate": 0.0005776950877868309,
"loss": 0.267,
"step": 52000
},
{
"epoch": 3.368203252814936,
"grad_norm": 1.1140522956848145,
"learning_rate": 0.0005755565535382049,
"loss": 0.2655,
"step": 52500
},
{
"epoch": 3.4002822955763,
"grad_norm": 1.4667320251464844,
"learning_rate": 0.0005734222963580762,
"loss": 0.2652,
"step": 53000
},
{
"epoch": 3.4006672440894365,
"eval_loss": 0.25107353925704956,
"eval_runtime": 8.1009,
"eval_samples_per_second": 61.721,
"eval_steps_per_second": 7.777,
"step": 53006
},
{
"epoch": 3.432361338337664,
"grad_norm": 1.0906621217727661,
"learning_rate": 0.0005712837621094502,
"loss": 0.2633,
"step": 53500
},
{
"epoch": 3.464440381099028,
"grad_norm": 1.7066080570220947,
"learning_rate": 0.0005691452278608242,
"loss": 0.2624,
"step": 54000
},
{
"epoch": 3.496519423860392,
"grad_norm": 1.1421736478805542,
"learning_rate": 0.0005670066936121983,
"loss": 0.267,
"step": 54500
},
{
"epoch": 3.528598466621756,
"grad_norm": 0.9186555743217468,
"learning_rate": 0.0005648681593635722,
"loss": 0.2646,
"step": 55000
},
{
"epoch": 3.5606775093831198,
"grad_norm": 0.8647829294204712,
"learning_rate": 0.0005627339021834436,
"loss": 0.2614,
"step": 55500
},
{
"epoch": 3.592756552144484,
"grad_norm": 0.9938694834709167,
"learning_rate": 0.0005605953679348175,
"loss": 0.263,
"step": 56000
},
{
"epoch": 3.6007121547493024,
"eval_loss": 0.2563716173171997,
"eval_runtime": 6.467,
"eval_samples_per_second": 77.315,
"eval_steps_per_second": 9.742,
"step": 56124
},
{
"epoch": 3.624835594905848,
"grad_norm": 0.8723744750022888,
"learning_rate": 0.0005584568336861915,
"loss": 0.2612,
"step": 56500
},
{
"epoch": 3.656914637667212,
"grad_norm": 1.0959194898605347,
"learning_rate": 0.0005563182994375655,
"loss": 0.2611,
"step": 57000
},
{
"epoch": 3.688993680428576,
"grad_norm": 1.2405571937561035,
"learning_rate": 0.0005541797651889395,
"loss": 0.2635,
"step": 57500
},
{
"epoch": 3.72107272318994,
"grad_norm": 1.137342929840088,
"learning_rate": 0.0005520412309403136,
"loss": 0.2595,
"step": 58000
},
{
"epoch": 3.753151765951304,
"grad_norm": 0.8755650520324707,
"learning_rate": 0.0005499026966916875,
"loss": 0.2619,
"step": 58500
},
{
"epoch": 3.785230808712668,
"grad_norm": 1.2968310117721558,
"learning_rate": 0.0005477641624430616,
"loss": 0.2612,
"step": 59000
},
{
"epoch": 3.8007570654091682,
"eval_loss": 0.2506495714187622,
"eval_runtime": 8.1115,
"eval_samples_per_second": 61.641,
"eval_steps_per_second": 7.767,
"step": 59242
},
{
"epoch": 3.817309851474032,
"grad_norm": 0.9842492938041687,
"learning_rate": 0.0005456299052629328,
"loss": 0.2624,
"step": 59500
},
{
"epoch": 3.849388894235396,
"grad_norm": 0.8897719979286194,
"learning_rate": 0.0005434913710143068,
"loss": 0.2573,
"step": 60000
},
{
"epoch": 3.88146793699676,
"grad_norm": 0.9902140498161316,
"learning_rate": 0.0005413528367656809,
"loss": 0.2624,
"step": 60500
},
{
"epoch": 3.913546979758124,
"grad_norm": 0.9460390210151672,
"learning_rate": 0.0005392185795855521,
"loss": 0.2582,
"step": 61000
},
{
"epoch": 3.945626022519488,
"grad_norm": 1.1403892040252686,
"learning_rate": 0.0005370800453369261,
"loss": 0.2627,
"step": 61500
},
{
"epoch": 3.977705065280852,
"grad_norm": 0.987301766872406,
"learning_rate": 0.0005349415110883001,
"loss": 0.2616,
"step": 62000
},
{
"epoch": 4.000769897026273,
"eval_loss": 0.2439550757408142,
"eval_runtime": 7.3235,
"eval_samples_per_second": 68.274,
"eval_steps_per_second": 8.602,
"step": 62360
},
{
"epoch": 4.009752028999455,
"grad_norm": 0.9634618163108826,
"learning_rate": 0.0005328029768396741,
"loss": 0.2522,
"step": 62500
},
{
"epoch": 4.041831071760819,
"grad_norm": 1.0599830150604248,
"learning_rate": 0.0005306687196595454,
"loss": 0.2548,
"step": 63000
},
{
"epoch": 4.0739101145221825,
"grad_norm": 1.1028203964233398,
"learning_rate": 0.0005285301854109194,
"loss": 0.2529,
"step": 63500
},
{
"epoch": 4.105989157283546,
"grad_norm": 1.1508251428604126,
"learning_rate": 0.0005263916511622933,
"loss": 0.2539,
"step": 64000
},
{
"epoch": 4.138068200044911,
"grad_norm": 1.0701543092727661,
"learning_rate": 0.0005242531169136675,
"loss": 0.2502,
"step": 64500
},
{
"epoch": 4.170147242806275,
"grad_norm": 1.0984095335006714,
"learning_rate": 0.0005221145826650414,
"loss": 0.2574,
"step": 65000
},
{
"epoch": 4.200814807686139,
"eval_loss": 0.24038133025169373,
"eval_runtime": 6.2912,
"eval_samples_per_second": 79.476,
"eval_steps_per_second": 10.014,
"step": 65478
},
{
"epoch": 4.202226285567638,
"grad_norm": 0.9908430576324463,
"learning_rate": 0.0005199803254849126,
"loss": 0.2564,
"step": 65500
},
{
"epoch": 4.234305328329003,
"grad_norm": 1.0797010660171509,
"learning_rate": 0.0005178417912362867,
"loss": 0.2568,
"step": 66000
},
{
"epoch": 4.266384371090367,
"grad_norm": 1.3110967874526978,
"learning_rate": 0.0005157032569876607,
"loss": 0.2546,
"step": 66500
},
{
"epoch": 4.298463413851731,
"grad_norm": 1.22694993019104,
"learning_rate": 0.0005135647227390347,
"loss": 0.253,
"step": 67000
},
{
"epoch": 4.330542456613094,
"grad_norm": 1.146044373512268,
"learning_rate": 0.000511430465558906,
"loss": 0.2529,
"step": 67500
},
{
"epoch": 4.362621499374459,
"grad_norm": 1.0950199365615845,
"learning_rate": 0.0005092919313102799,
"loss": 0.2506,
"step": 68000
},
{
"epoch": 4.394700542135823,
"grad_norm": 1.1739941835403442,
"learning_rate": 0.0005071533970616539,
"loss": 0.2501,
"step": 68500
},
{
"epoch": 4.4008597183460045,
"eval_loss": 0.24005259573459625,
"eval_runtime": 7.6188,
"eval_samples_per_second": 65.627,
"eval_steps_per_second": 8.269,
"step": 68596
},
{
"epoch": 4.4267795848971865,
"grad_norm": 1.3007426261901855,
"learning_rate": 0.000505014862813028,
"loss": 0.2507,
"step": 69000
},
{
"epoch": 4.45885862765855,
"grad_norm": 1.0202151536941528,
"learning_rate": 0.0005028763285644019,
"loss": 0.2535,
"step": 69500
},
{
"epoch": 4.490937670419915,
"grad_norm": 1.5328317880630493,
"learning_rate": 0.000500737794315776,
"loss": 0.2522,
"step": 70000
},
{
"epoch": 4.523016713181279,
"grad_norm": 1.0239914655685425,
"learning_rate": 0.0004986035371356472,
"loss": 0.2496,
"step": 70500
},
{
"epoch": 4.555095755942642,
"grad_norm": 0.9700740575790405,
"learning_rate": 0.0004964650028870213,
"loss": 0.2512,
"step": 71000
},
{
"epoch": 4.587174798704007,
"grad_norm": 1.2554644346237183,
"learning_rate": 0.0004943264686383952,
"loss": 0.2506,
"step": 71500
},
{
"epoch": 4.60090462900587,
"eval_loss": 0.24138595163822174,
"eval_runtime": 8.1086,
"eval_samples_per_second": 61.663,
"eval_steps_per_second": 7.77,
"step": 71714
},
{
"epoch": 4.619253841465371,
"grad_norm": 1.5016344785690308,
"learning_rate": 0.0004921879343897693,
"loss": 0.2471,
"step": 72000
},
{
"epoch": 4.6513328842267345,
"grad_norm": 1.251976490020752,
"learning_rate": 0.0004900494001411433,
"loss": 0.2531,
"step": 72500
},
{
"epoch": 4.683411926988098,
"grad_norm": 1.0139933824539185,
"learning_rate": 0.0004879151429610145,
"loss": 0.2488,
"step": 73000
},
{
"epoch": 4.715490969749463,
"grad_norm": 0.9058252573013306,
"learning_rate": 0.0004857766087123886,
"loss": 0.2478,
"step": 73500
},
{
"epoch": 4.747570012510827,
"grad_norm": 0.8362458944320679,
"learning_rate": 0.00048363807446376257,
"loss": 0.2495,
"step": 74000
},
{
"epoch": 4.77964905527219,
"grad_norm": 0.996514081954956,
"learning_rate": 0.00048149954021513655,
"loss": 0.251,
"step": 74500
},
{
"epoch": 4.800949539665736,
"eval_loss": 0.23780158162117004,
"eval_runtime": 6.3001,
"eval_samples_per_second": 79.364,
"eval_steps_per_second": 10.0,
"step": 74832
},
{
"epoch": 4.811728098033555,
"grad_norm": 1.115113377571106,
"learning_rate": 0.00047936528303500785,
"loss": 0.2477,
"step": 75000
},
{
"epoch": 4.843807140794919,
"grad_norm": 1.2068469524383545,
"learning_rate": 0.00047722674878638183,
"loss": 0.2462,
"step": 75500
},
{
"epoch": 4.875886183556283,
"grad_norm": 1.0453870296478271,
"learning_rate": 0.0004750882145377558,
"loss": 0.2456,
"step": 76000
},
{
"epoch": 4.907965226317646,
"grad_norm": 1.0999488830566406,
"learning_rate": 0.0004729496802891299,
"loss": 0.2492,
"step": 76500
},
{
"epoch": 4.940044269079011,
"grad_norm": 0.8874688148498535,
"learning_rate": 0.00047081114604050387,
"loss": 0.2467,
"step": 77000
},
{
"epoch": 4.972123311840375,
"grad_norm": 1.060391902923584,
"learning_rate": 0.0004686726117918779,
"loss": 0.2456,
"step": 77500
}
],
"logging_steps": 500,
"max_steps": 187044,
"num_input_tokens_seen": 0,
"num_train_epochs": 12,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.2869610745448366e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}