Safetensors
CosFly-Track / checkpoints /Qwen3-VL-2B-SFT /trainer_state.json
renruilong's picture
Add scripts and checkpoints (CosFly-Track release) (#85)
284878d
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 3125,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0032,
"grad_norm": 16.167704755253098,
"learning_rate": 1.437699680511182e-07,
"loss": 0.6528051853179931,
"step": 10
},
{
"epoch": 0.0064,
"grad_norm": 15.890120546753822,
"learning_rate": 3.0351437699680514e-07,
"loss": 0.6462714195251464,
"step": 20
},
{
"epoch": 0.0096,
"grad_norm": 14.94996510180698,
"learning_rate": 4.6325878594249205e-07,
"loss": 0.6038930416107178,
"step": 30
},
{
"epoch": 0.0128,
"grad_norm": 7.595956825837255,
"learning_rate": 6.230031948881789e-07,
"loss": 0.49077792167663575,
"step": 40
},
{
"epoch": 0.016,
"grad_norm": 3.026643067758099,
"learning_rate": 7.82747603833866e-07,
"loss": 0.3725566864013672,
"step": 50
},
{
"epoch": 0.0192,
"grad_norm": 1.45050871801394,
"learning_rate": 9.424920127795528e-07,
"loss": 0.3130798816680908,
"step": 60
},
{
"epoch": 0.0224,
"grad_norm": 0.7098603642718405,
"learning_rate": 1.1022364217252397e-06,
"loss": 0.29621334075927735,
"step": 70
},
{
"epoch": 0.0256,
"grad_norm": 0.6027577608327673,
"learning_rate": 1.2619808306709266e-06,
"loss": 0.27455599308013917,
"step": 80
},
{
"epoch": 0.0288,
"grad_norm": 0.6521596145147045,
"learning_rate": 1.4217252396166134e-06,
"loss": 0.2667043447494507,
"step": 90
},
{
"epoch": 0.032,
"grad_norm": 0.5069890685833461,
"learning_rate": 1.5814696485623005e-06,
"loss": 0.26807360649108886,
"step": 100
},
{
"epoch": 0.0352,
"grad_norm": 0.5470393023746721,
"learning_rate": 1.7412140575079875e-06,
"loss": 0.26680865287780764,
"step": 110
},
{
"epoch": 0.0384,
"grad_norm": 0.5543553869620175,
"learning_rate": 1.9009584664536742e-06,
"loss": 0.25434055328369143,
"step": 120
},
{
"epoch": 0.0416,
"grad_norm": 0.5420531484574165,
"learning_rate": 2.060702875399361e-06,
"loss": 0.25767529010772705,
"step": 130
},
{
"epoch": 0.0448,
"grad_norm": 0.645702037816744,
"learning_rate": 2.220447284345048e-06,
"loss": 0.24863953590393068,
"step": 140
},
{
"epoch": 0.048,
"grad_norm": 0.6143136416629473,
"learning_rate": 2.380191693290735e-06,
"loss": 0.24553947448730468,
"step": 150
},
{
"epoch": 0.0512,
"grad_norm": 0.5094817219127052,
"learning_rate": 2.539936102236422e-06,
"loss": 0.2415369987487793,
"step": 160
},
{
"epoch": 0.0544,
"grad_norm": 0.6291606522275387,
"learning_rate": 2.699680511182109e-06,
"loss": 0.24887418746948242,
"step": 170
},
{
"epoch": 0.0576,
"grad_norm": 0.6248895072998087,
"learning_rate": 2.8594249201277955e-06,
"loss": 0.2414403438568115,
"step": 180
},
{
"epoch": 0.0608,
"grad_norm": 0.6640745861299296,
"learning_rate": 3.0191693290734825e-06,
"loss": 0.24553894996643066,
"step": 190
},
{
"epoch": 0.064,
"grad_norm": 0.6136916428260776,
"learning_rate": 3.17891373801917e-06,
"loss": 0.24655485153198242,
"step": 200
},
{
"epoch": 0.0672,
"grad_norm": 0.6572881584027297,
"learning_rate": 3.3386581469648564e-06,
"loss": 0.2433255910873413,
"step": 210
},
{
"epoch": 0.0704,
"grad_norm": 0.6365580690264084,
"learning_rate": 3.4984025559105434e-06,
"loss": 0.23687341213226318,
"step": 220
},
{
"epoch": 0.0736,
"grad_norm": 0.6771736107097397,
"learning_rate": 3.6581469648562303e-06,
"loss": 0.23829469680786133,
"step": 230
},
{
"epoch": 0.0768,
"grad_norm": 0.6990706788858505,
"learning_rate": 3.817891373801918e-06,
"loss": 0.23471264839172362,
"step": 240
},
{
"epoch": 0.08,
"grad_norm": 0.6029376877872676,
"learning_rate": 3.977635782747604e-06,
"loss": 0.23215394020080565,
"step": 250
},
{
"epoch": 0.0832,
"grad_norm": 0.6082124769869354,
"learning_rate": 4.137380191693291e-06,
"loss": 0.2326298713684082,
"step": 260
},
{
"epoch": 0.0864,
"grad_norm": 0.7069824323872274,
"learning_rate": 4.297124600638978e-06,
"loss": 0.23525137901306153,
"step": 270
},
{
"epoch": 0.0896,
"grad_norm": 0.6697633994539672,
"learning_rate": 4.456869009584665e-06,
"loss": 0.23122966289520264,
"step": 280
},
{
"epoch": 0.0928,
"grad_norm": 0.5896144959913211,
"learning_rate": 4.616613418530352e-06,
"loss": 0.2369994878768921,
"step": 290
},
{
"epoch": 0.096,
"grad_norm": 0.6202443536122002,
"learning_rate": 4.776357827476039e-06,
"loss": 0.23878774642944336,
"step": 300
},
{
"epoch": 0.0992,
"grad_norm": 0.654740818437731,
"learning_rate": 4.936102236421725e-06,
"loss": 0.22523627281188965,
"step": 310
},
{
"epoch": 0.1024,
"grad_norm": 0.5332231058888761,
"learning_rate": 4.999943833158769e-06,
"loss": 0.22634780406951904,
"step": 320
},
{
"epoch": 0.1056,
"grad_norm": 0.5353007164619794,
"learning_rate": 4.999600600490783e-06,
"loss": 0.23276047706604003,
"step": 330
},
{
"epoch": 0.1088,
"grad_norm": 0.53617134295571,
"learning_rate": 4.9989453817439345e-06,
"loss": 0.22672569751739502,
"step": 340
},
{
"epoch": 0.112,
"grad_norm": 0.5149149938648103,
"learning_rate": 4.997978258698942e-06,
"loss": 0.22631363868713378,
"step": 350
},
{
"epoch": 0.1152,
"grad_norm": 0.5959881018141326,
"learning_rate": 4.996699352066659e-06,
"loss": 0.22707018852233887,
"step": 360
},
{
"epoch": 0.1184,
"grad_norm": 0.6648028246958526,
"learning_rate": 4.995108821473014e-06,
"loss": 0.22777373790740968,
"step": 370
},
{
"epoch": 0.1216,
"grad_norm": 0.6395047869916185,
"learning_rate": 4.993206865439084e-06,
"loss": 0.22382116317749023,
"step": 380
},
{
"epoch": 0.1248,
"grad_norm": 0.6449783716947614,
"learning_rate": 4.990993721356317e-06,
"loss": 0.22268824577331542,
"step": 390
},
{
"epoch": 0.128,
"grad_norm": 0.6709421623745665,
"learning_rate": 4.988469665456901e-06,
"loss": 0.22317943572998047,
"step": 400
},
{
"epoch": 0.1312,
"grad_norm": 0.5466948727484514,
"learning_rate": 4.985635012779288e-06,
"loss": 0.23101482391357422,
"step": 410
},
{
"epoch": 0.1344,
"grad_norm": 0.48989327197226856,
"learning_rate": 4.98249011712887e-06,
"loss": 0.2234072208404541,
"step": 420
},
{
"epoch": 0.1376,
"grad_norm": 0.5417400145938138,
"learning_rate": 4.979035371033824e-06,
"loss": 0.22212049961090088,
"step": 430
},
{
"epoch": 0.1408,
"grad_norm": 0.5576422767413268,
"learning_rate": 4.975271205696115e-06,
"loss": 0.22228083610534669,
"step": 440
},
{
"epoch": 0.144,
"grad_norm": 0.6175584799790863,
"learning_rate": 4.971198090937671e-06,
"loss": 0.21532373428344725,
"step": 450
},
{
"epoch": 0.1472,
"grad_norm": 0.6360712146764758,
"learning_rate": 4.966816535141756e-06,
"loss": 0.21311187744140625,
"step": 460
},
{
"epoch": 0.1504,
"grad_norm": 0.5401953881204377,
"learning_rate": 4.9621270851895035e-06,
"loss": 0.22237277030944824,
"step": 470
},
{
"epoch": 0.1536,
"grad_norm": 0.5988873649948656,
"learning_rate": 4.957130326391662e-06,
"loss": 0.22391064167022706,
"step": 480
},
{
"epoch": 0.1568,
"grad_norm": 0.5132670412160366,
"learning_rate": 4.951826882415544e-06,
"loss": 0.2206397533416748,
"step": 490
},
{
"epoch": 0.16,
"grad_norm": 0.5935020011592513,
"learning_rate": 4.946217415207177e-06,
"loss": 0.2148068904876709,
"step": 500
},
{
"epoch": 0.1632,
"grad_norm": 0.5324390349507315,
"learning_rate": 4.940302624908689e-06,
"loss": 0.21909193992614745,
"step": 510
},
{
"epoch": 0.1664,
"grad_norm": 0.6082929578051663,
"learning_rate": 4.934083249770912e-06,
"loss": 0.2133782386779785,
"step": 520
},
{
"epoch": 0.1696,
"grad_norm": 0.6272295187969801,
"learning_rate": 4.927560066061251e-06,
"loss": 0.2180723190307617,
"step": 530
},
{
"epoch": 0.1728,
"grad_norm": 0.5538741111929965,
"learning_rate": 4.920733887966783e-06,
"loss": 0.22759020328521729,
"step": 540
},
{
"epoch": 0.176,
"grad_norm": 0.5703593568416581,
"learning_rate": 4.913605567492636e-06,
"loss": 0.21657073497772217,
"step": 550
},
{
"epoch": 0.1792,
"grad_norm": 0.5873043850881617,
"learning_rate": 4.906175994355656e-06,
"loss": 0.21824207305908203,
"step": 560
},
{
"epoch": 0.1824,
"grad_norm": 0.7955355117519857,
"learning_rate": 4.898446095873345e-06,
"loss": 0.2209712028503418,
"step": 570
},
{
"epoch": 0.1856,
"grad_norm": 0.5347403539894492,
"learning_rate": 4.890416836848128e-06,
"loss": 0.2184591293334961,
"step": 580
},
{
"epoch": 0.1888,
"grad_norm": 0.5464598874722423,
"learning_rate": 4.882089219446925e-06,
"loss": 0.2130581855773926,
"step": 590
},
{
"epoch": 0.192,
"grad_norm": 0.5871382794412585,
"learning_rate": 4.873464283076074e-06,
"loss": 0.21770844459533692,
"step": 600
},
{
"epoch": 0.1952,
"grad_norm": 0.5516595084585112,
"learning_rate": 4.864543104251587e-06,
"loss": 0.21629047393798828,
"step": 610
},
{
"epoch": 0.1984,
"grad_norm": 0.5949100146178041,
"learning_rate": 4.855326796464798e-06,
"loss": 0.22033746242523194,
"step": 620
},
{
"epoch": 0.2016,
"grad_norm": 0.5798876425998256,
"learning_rate": 4.8458165100433725e-06,
"loss": 0.21477458477020264,
"step": 630
},
{
"epoch": 0.2048,
"grad_norm": 0.563545251458103,
"learning_rate": 4.836013432007738e-06,
"loss": 0.21490144729614258,
"step": 640
},
{
"epoch": 0.208,
"grad_norm": 0.5256728978801903,
"learning_rate": 4.825918785922921e-06,
"loss": 0.21858677864074708,
"step": 650
},
{
"epoch": 0.2112,
"grad_norm": 0.5062609806869888,
"learning_rate": 4.8155338317458315e-06,
"loss": 0.21592459678649903,
"step": 660
},
{
"epoch": 0.2144,
"grad_norm": 0.555318042395406,
"learning_rate": 4.804859865668002e-06,
"loss": 0.21323423385620116,
"step": 670
},
{
"epoch": 0.2176,
"grad_norm": 0.6382467151310525,
"learning_rate": 4.793898219953804e-06,
"loss": 0.2151188373565674,
"step": 680
},
{
"epoch": 0.2208,
"grad_norm": 0.5426280956852546,
"learning_rate": 4.782650262774164e-06,
"loss": 0.2155141830444336,
"step": 690
},
{
"epoch": 0.224,
"grad_norm": 0.5602841392771764,
"learning_rate": 4.7711173980357886e-06,
"loss": 0.21242978572845458,
"step": 700
},
{
"epoch": 0.2272,
"grad_norm": 0.5837827171492797,
"learning_rate": 4.759301065205947e-06,
"loss": 0.2129213333129883,
"step": 710
},
{
"epoch": 0.2304,
"grad_norm": 0.5678516858648391,
"learning_rate": 4.7472027391328e-06,
"loss": 0.21422340869903564,
"step": 720
},
{
"epoch": 0.2336,
"grad_norm": 0.6213695156464779,
"learning_rate": 4.734823929861317e-06,
"loss": 0.2172607660293579,
"step": 730
},
{
"epoch": 0.2368,
"grad_norm": 0.6084105321286742,
"learning_rate": 4.722166182444801e-06,
"loss": 0.21331138610839845,
"step": 740
},
{
"epoch": 0.24,
"grad_norm": 0.5848312022835148,
"learning_rate": 4.709231076752045e-06,
"loss": 0.21255254745483398,
"step": 750
},
{
"epoch": 0.2432,
"grad_norm": 0.5855428740644943,
"learning_rate": 4.696020227270142e-06,
"loss": 0.21734881401062012,
"step": 760
},
{
"epoch": 0.2464,
"grad_norm": 0.5135013968609298,
"learning_rate": 4.6825352829029705e-06,
"loss": 0.21321442127227783,
"step": 770
},
{
"epoch": 0.2496,
"grad_norm": 0.5938685951597557,
"learning_rate": 4.668777926765392e-06,
"loss": 0.21113758087158202,
"step": 780
},
{
"epoch": 0.2528,
"grad_norm": 0.6490004462160337,
"learning_rate": 4.6547498759731725e-06,
"loss": 0.20692987442016603,
"step": 790
},
{
"epoch": 0.256,
"grad_norm": 0.5694207965471786,
"learning_rate": 4.6404528814286575e-06,
"loss": 0.20959222316741943,
"step": 800
},
{
"epoch": 0.2592,
"grad_norm": 0.5648942925010132,
"learning_rate": 4.6258887276022425e-06,
"loss": 0.21758944988250734,
"step": 810
},
{
"epoch": 0.2624,
"grad_norm": 0.6544068998265237,
"learning_rate": 4.611059232309639e-06,
"loss": 0.21146907806396484,
"step": 820
},
{
"epoch": 0.2656,
"grad_norm": 0.6680185905090128,
"learning_rate": 4.595966246484986e-06,
"loss": 0.21348462104797364,
"step": 830
},
{
"epoch": 0.2688,
"grad_norm": 0.4956164506371995,
"learning_rate": 4.580611653949829e-06,
"loss": 0.21317172050476074,
"step": 840
},
{
"epoch": 0.272,
"grad_norm": 0.6491508776235345,
"learning_rate": 4.564997371177992e-06,
"loss": 0.2108323574066162,
"step": 850
},
{
"epoch": 0.2752,
"grad_norm": 0.6859739128419746,
"learning_rate": 4.54912534705637e-06,
"loss": 0.21068863868713378,
"step": 860
},
{
"epoch": 0.2784,
"grad_norm": 0.5876140035889241,
"learning_rate": 4.532997562641683e-06,
"loss": 0.20738301277160645,
"step": 870
},
{
"epoch": 0.2816,
"grad_norm": 0.5388630641864397,
"learning_rate": 4.516616030913214e-06,
"loss": 0.2113194465637207,
"step": 880
},
{
"epoch": 0.2848,
"grad_norm": 0.527263546069221,
"learning_rate": 4.499982796521556e-06,
"loss": 0.20718231201171874,
"step": 890
},
{
"epoch": 0.288,
"grad_norm": 0.6778383199902553,
"learning_rate": 4.48309993553341e-06,
"loss": 0.20899975299835205,
"step": 900
},
{
"epoch": 0.2912,
"grad_norm": 0.6041502046582736,
"learning_rate": 4.465969555172468e-06,
"loss": 0.20922982692718506,
"step": 910
},
{
"epoch": 0.2944,
"grad_norm": 0.5872507915529911,
"learning_rate": 4.448593793556391e-06,
"loss": 0.21518073081970215,
"step": 920
},
{
"epoch": 0.2976,
"grad_norm": 0.5414243473578003,
"learning_rate": 4.430974819429954e-06,
"loss": 0.20869126319885253,
"step": 930
},
{
"epoch": 0.3008,
"grad_norm": 0.4624854855413159,
"learning_rate": 4.413114831894344e-06,
"loss": 0.20277881622314453,
"step": 940
},
{
"epoch": 0.304,
"grad_norm": 0.5247854876993729,
"learning_rate": 4.3950160601326865e-06,
"loss": 0.20181698799133302,
"step": 950
},
{
"epoch": 0.3072,
"grad_norm": 0.5808078368512252,
"learning_rate": 4.376680763131811e-06,
"loss": 0.20898809432983398,
"step": 960
},
{
"epoch": 0.3104,
"grad_norm": 0.5805212694083882,
"learning_rate": 4.358111229400296e-06,
"loss": 0.21212198734283447,
"step": 970
},
{
"epoch": 0.3136,
"grad_norm": 0.5721764020420262,
"learning_rate": 4.33930977668283e-06,
"loss": 0.21448736190795897,
"step": 980
},
{
"epoch": 0.3168,
"grad_norm": 0.5598008397585128,
"learning_rate": 4.320278751670922e-06,
"loss": 0.20758256912231446,
"step": 990
},
{
"epoch": 0.32,
"grad_norm": 0.5522723710696453,
"learning_rate": 4.301020529710009e-06,
"loss": 0.20947573184967042,
"step": 1000
},
{
"epoch": 0.3232,
"grad_norm": 0.5556932215476815,
"learning_rate": 4.281537514502962e-06,
"loss": 0.2131945848464966,
"step": 1010
},
{
"epoch": 0.3264,
"grad_norm": 0.5256326530235461,
"learning_rate": 4.261832137810093e-06,
"loss": 0.20962438583374024,
"step": 1020
},
{
"epoch": 0.3296,
"grad_norm": 0.5141067804644184,
"learning_rate": 4.241906859145611e-06,
"loss": 0.21035046577453614,
"step": 1030
},
{
"epoch": 0.3328,
"grad_norm": 0.509376911595103,
"learning_rate": 4.221764165470661e-06,
"loss": 0.20757730007171632,
"step": 1040
},
{
"epoch": 0.336,
"grad_norm": 0.5632632187185198,
"learning_rate": 4.201406570882898e-06,
"loss": 0.20691304206848143,
"step": 1050
},
{
"epoch": 0.3392,
"grad_norm": 0.5786515758035645,
"learning_rate": 4.180836616302704e-06,
"loss": 0.20582923889160157,
"step": 1060
},
{
"epoch": 0.3424,
"grad_norm": 0.591108109764431,
"learning_rate": 4.160056869156041e-06,
"loss": 0.2102893590927124,
"step": 1070
},
{
"epoch": 0.3456,
"grad_norm": 0.5367428274966828,
"learning_rate": 4.139069923053995e-06,
"loss": 0.20834057331085204,
"step": 1080
},
{
"epoch": 0.3488,
"grad_norm": 0.49962583382458753,
"learning_rate": 4.117878397469062e-06,
"loss": 0.2114588975906372,
"step": 1090
},
{
"epoch": 0.352,
"grad_norm": 0.5580828852277292,
"learning_rate": 4.096484937408195e-06,
"loss": 0.2029412269592285,
"step": 1100
},
{
"epoch": 0.3552,
"grad_norm": 0.5671943339841842,
"learning_rate": 4.074892213082676e-06,
"loss": 0.20308828353881836,
"step": 1110
},
{
"epoch": 0.3584,
"grad_norm": 0.5583868175031171,
"learning_rate": 4.0531029195748265e-06,
"loss": 0.2104210376739502,
"step": 1120
},
{
"epoch": 0.3616,
"grad_norm": 0.5452939479895703,
"learning_rate": 4.03111977650163e-06,
"loss": 0.20968456268310548,
"step": 1130
},
{
"epoch": 0.3648,
"grad_norm": 0.6195183591357212,
"learning_rate": 4.008945527675281e-06,
"loss": 0.20957679748535157,
"step": 1140
},
{
"epoch": 0.368,
"grad_norm": 0.6171258889408775,
"learning_rate": 3.986582940760717e-06,
"loss": 0.1984492540359497,
"step": 1150
},
{
"epoch": 0.3712,
"grad_norm": 0.6164010362674036,
"learning_rate": 3.9640348069301785e-06,
"loss": 0.20632429122924806,
"step": 1160
},
{
"epoch": 0.3744,
"grad_norm": 0.5558070727772452,
"learning_rate": 3.941303940514826e-06,
"loss": 0.20776019096374512,
"step": 1170
},
{
"epoch": 0.3776,
"grad_norm": 0.5943916453083408,
"learning_rate": 3.918393178653472e-06,
"loss": 0.20839078426361085,
"step": 1180
},
{
"epoch": 0.3808,
"grad_norm": 0.5018385923371635,
"learning_rate": 3.895305380938468e-06,
"loss": 0.2044908285140991,
"step": 1190
},
{
"epoch": 0.384,
"grad_norm": 0.48660847876218716,
"learning_rate": 3.872043429058783e-06,
"loss": 0.20328717231750487,
"step": 1200
},
{
"epoch": 0.3872,
"grad_norm": 0.5586353975354608,
"learning_rate": 3.84861022644033e-06,
"loss": 0.20572426319122314,
"step": 1210
},
{
"epoch": 0.3904,
"grad_norm": 0.5709168788921625,
"learning_rate": 3.825008697883574e-06,
"loss": 0.21369614601135253,
"step": 1220
},
{
"epoch": 0.3936,
"grad_norm": 0.5589246090839964,
"learning_rate": 3.8012417891984776e-06,
"loss": 0.2072831630706787,
"step": 1230
},
{
"epoch": 0.3968,
"grad_norm": 0.5711782327133378,
"learning_rate": 3.777312466836819e-06,
"loss": 0.20526669025421143,
"step": 1240
},
{
"epoch": 0.4,
"grad_norm": 0.5656399244912672,
"learning_rate": 3.7532237175219378e-06,
"loss": 0.20442888736724854,
"step": 1250
},
{
"epoch": 0.4032,
"grad_norm": 0.5520901024337347,
"learning_rate": 3.728978547875948e-06,
"loss": 0.2092284679412842,
"step": 1260
},
{
"epoch": 0.4064,
"grad_norm": 0.553756025103199,
"learning_rate": 3.7045799840444712e-06,
"loss": 0.20277605056762696,
"step": 1270
},
{
"epoch": 0.4096,
"grad_norm": 0.5430187148138641,
"learning_rate": 3.6800310713189258e-06,
"loss": 0.20491743087768555,
"step": 1280
},
{
"epoch": 0.4128,
"grad_norm": 0.7620941398223869,
"learning_rate": 3.6553348737564328e-06,
"loss": 0.2055516481399536,
"step": 1290
},
{
"epoch": 0.416,
"grad_norm": 0.5265612798122297,
"learning_rate": 3.6304944737973794e-06,
"loss": 0.21130599975585937,
"step": 1300
},
{
"epoch": 0.4192,
"grad_norm": 0.5353794185025008,
"learning_rate": 3.6055129718806836e-06,
"loss": 0.20504627227783204,
"step": 1310
},
{
"epoch": 0.4224,
"grad_norm": 0.5979654766960453,
"learning_rate": 3.5803934860568134e-06,
"loss": 0.2000981330871582,
"step": 1320
},
{
"epoch": 0.4256,
"grad_norm": 0.5915664314356317,
"learning_rate": 3.5551391515986163e-06,
"loss": 0.20581989288330077,
"step": 1330
},
{
"epoch": 0.4288,
"grad_norm": 0.562992516341074,
"learning_rate": 3.529753120609982e-06,
"loss": 0.20160207748413086,
"step": 1340
},
{
"epoch": 0.432,
"grad_norm": 0.7046032478558245,
"learning_rate": 3.5042385616324243e-06,
"loss": 0.2043483018875122,
"step": 1350
},
{
"epoch": 0.4352,
"grad_norm": 0.5184492477363449,
"learning_rate": 3.4785986592495934e-06,
"loss": 0.20285494327545167,
"step": 1360
},
{
"epoch": 0.4384,
"grad_norm": 0.5806380074338086,
"learning_rate": 3.452836613689803e-06,
"loss": 0.2009434223175049,
"step": 1370
},
{
"epoch": 0.4416,
"grad_norm": 0.5204618736945451,
"learning_rate": 3.426955640426584e-06,
"loss": 0.20416510105133057,
"step": 1380
},
{
"epoch": 0.4448,
"grad_norm": 0.5765864502605341,
"learning_rate": 3.4009589697773605e-06,
"loss": 0.20326631069183348,
"step": 1390
},
{
"epoch": 0.448,
"grad_norm": 0.5779970501460372,
"learning_rate": 3.3748498465002475e-06,
"loss": 0.20073289871215821,
"step": 1400
},
{
"epoch": 0.4512,
"grad_norm": 0.6393995362823897,
"learning_rate": 3.3486315293890693e-06,
"loss": 0.20874643325805664,
"step": 1410
},
{
"epoch": 0.4544,
"grad_norm": 0.5108762095593324,
"learning_rate": 3.3223072908666053e-06,
"loss": 0.19835340976715088,
"step": 1420
},
{
"epoch": 0.4576,
"grad_norm": 0.6435280387445825,
"learning_rate": 3.295880416576153e-06,
"loss": 0.20992684364318848,
"step": 1430
},
{
"epoch": 0.4608,
"grad_norm": 0.5838753206875198,
"learning_rate": 3.269354204971427e-06,
"loss": 0.20265870094299315,
"step": 1440
},
{
"epoch": 0.464,
"grad_norm": 0.6745984788898958,
"learning_rate": 3.242731966904865e-06,
"loss": 0.20037527084350587,
"step": 1450
},
{
"epoch": 0.4672,
"grad_norm": 0.5358161645108944,
"learning_rate": 3.2160170252143913e-06,
"loss": 0.20123369693756105,
"step": 1460
},
{
"epoch": 0.4704,
"grad_norm": 0.5112361606823973,
"learning_rate": 3.1892127143086716e-06,
"loss": 0.20752406120300293,
"step": 1470
},
{
"epoch": 0.4736,
"grad_norm": 0.6333759965752455,
"learning_rate": 3.1623223797509347e-06,
"loss": 0.19706425666809083,
"step": 1480
},
{
"epoch": 0.4768,
"grad_norm": 0.6206117536462172,
"learning_rate": 3.135349377841396e-06,
"loss": 0.20125732421875,
"step": 1490
},
{
"epoch": 0.48,
"grad_norm": 0.5541712474486513,
"learning_rate": 3.1082970751983497e-06,
"loss": 0.20749812126159667,
"step": 1500
},
{
"epoch": 0.4832,
"grad_norm": 0.5835934183180771,
"learning_rate": 3.0811688483379546e-06,
"loss": 0.20475554466247559,
"step": 1510
},
{
"epoch": 0.4864,
"grad_norm": 0.5792514427898341,
"learning_rate": 3.0539680832528074e-06,
"loss": 0.20504088401794435,
"step": 1520
},
{
"epoch": 0.4896,
"grad_norm": 0.6358843481166787,
"learning_rate": 3.026698174989316e-06,
"loss": 0.20325000286102296,
"step": 1530
},
{
"epoch": 0.4928,
"grad_norm": 0.5059500889981753,
"learning_rate": 2.999362527223952e-06,
"loss": 0.2031909465789795,
"step": 1540
},
{
"epoch": 0.496,
"grad_norm": 0.5388306821924389,
"learning_rate": 2.9719645518384194e-06,
"loss": 0.20504312515258788,
"step": 1550
},
{
"epoch": 0.4992,
"grad_norm": 0.5939936480408617,
"learning_rate": 2.944507668493807e-06,
"loss": 0.2084404706954956,
"step": 1560
},
{
"epoch": 0.5024,
"grad_norm": 0.5687025114161597,
"learning_rate": 2.9169953042037623e-06,
"loss": 0.20367155075073243,
"step": 1570
},
{
"epoch": 0.5056,
"grad_norm": 0.5703613797457775,
"learning_rate": 2.889430892906754e-06,
"loss": 0.19950419664382935,
"step": 1580
},
{
"epoch": 0.5088,
"grad_norm": 0.50147360976836,
"learning_rate": 2.861817875037462e-06,
"loss": 0.19737675189971923,
"step": 1590
},
{
"epoch": 0.512,
"grad_norm": 0.5962810686359508,
"learning_rate": 2.8341596970973683e-06,
"loss": 0.206866455078125,
"step": 1600
},
{
"epoch": 0.5152,
"grad_norm": 0.564566320219468,
"learning_rate": 2.80645981122458e-06,
"loss": 0.2020205020904541,
"step": 1610
},
{
"epoch": 0.5184,
"grad_norm": 0.5246372929237232,
"learning_rate": 2.7787216747629508e-06,
"loss": 0.20939722061157226,
"step": 1620
},
{
"epoch": 0.5216,
"grad_norm": 0.5415181940486332,
"learning_rate": 2.7509487498305615e-06,
"loss": 0.19629446268081666,
"step": 1630
},
{
"epoch": 0.5248,
"grad_norm": 0.5627430222118958,
"learning_rate": 2.7231445028875924e-06,
"loss": 0.20240178108215331,
"step": 1640
},
{
"epoch": 0.528,
"grad_norm": 0.5578941065241574,
"learning_rate": 2.6953124043036604e-06,
"loss": 0.2012562036514282,
"step": 1650
},
{
"epoch": 0.5312,
"grad_norm": 0.5487117054063715,
"learning_rate": 2.667455927924667e-06,
"loss": 0.20127537250518798,
"step": 1660
},
{
"epoch": 0.5344,
"grad_norm": 0.571360126804376,
"learning_rate": 2.6395785506392164e-06,
"loss": 0.1964709758758545,
"step": 1670
},
{
"epoch": 0.5376,
"grad_norm": 0.6088527341362128,
"learning_rate": 2.6116837519446407e-06,
"loss": 0.1997244954109192,
"step": 1680
},
{
"epoch": 0.5408,
"grad_norm": 0.5974545138027041,
"learning_rate": 2.5837750135127192e-06,
"loss": 0.19768773317337035,
"step": 1690
},
{
"epoch": 0.544,
"grad_norm": 0.5496714163583045,
"learning_rate": 2.555855818755108e-06,
"loss": 0.20294923782348634,
"step": 1700
},
{
"epoch": 0.5472,
"grad_norm": 0.7083231030411815,
"learning_rate": 2.5279296523885636e-06,
"loss": 0.20083847045898437,
"step": 1710
},
{
"epoch": 0.5504,
"grad_norm": 0.5938882026412365,
"learning_rate": 2.5e-06,
"loss": 0.20156488418579102,
"step": 1720
},
{
"epoch": 0.5536,
"grad_norm": 0.5963429209905415,
"learning_rate": 2.472070347611437e-06,
"loss": 0.19514652490615844,
"step": 1730
},
{
"epoch": 0.5568,
"grad_norm": 0.6395947365412442,
"learning_rate": 2.444144181244893e-06,
"loss": 0.20121583938598633,
"step": 1740
},
{
"epoch": 0.56,
"grad_norm": 0.5998001248295249,
"learning_rate": 2.416224986487282e-06,
"loss": 0.19726226329803467,
"step": 1750
},
{
"epoch": 0.5632,
"grad_norm": 0.5593754591530539,
"learning_rate": 2.3883162480553605e-06,
"loss": 0.19497768878936766,
"step": 1760
},
{
"epoch": 0.5664,
"grad_norm": 0.5860785466160793,
"learning_rate": 2.3604214493607844e-06,
"loss": 0.1996150493621826,
"step": 1770
},
{
"epoch": 0.5696,
"grad_norm": 0.5963601131944923,
"learning_rate": 2.332544072075333e-06,
"loss": 0.20348951816558838,
"step": 1780
},
{
"epoch": 0.5728,
"grad_norm": 0.5745583695919886,
"learning_rate": 2.30468759569634e-06,
"loss": 0.2016512393951416,
"step": 1790
},
{
"epoch": 0.576,
"grad_norm": 0.5720738010975994,
"learning_rate": 2.276855497112408e-06,
"loss": 0.1983588457107544,
"step": 1800
},
{
"epoch": 0.5792,
"grad_norm": 0.594436652050367,
"learning_rate": 2.2490512501694394e-06,
"loss": 0.19393882751464844,
"step": 1810
},
{
"epoch": 0.5824,
"grad_norm": 0.5547702774883363,
"learning_rate": 2.2212783252370496e-06,
"loss": 0.19950855970382692,
"step": 1820
},
{
"epoch": 0.5856,
"grad_norm": 0.49741997333090354,
"learning_rate": 2.1935401887754213e-06,
"loss": 0.20486598014831542,
"step": 1830
},
{
"epoch": 0.5888,
"grad_norm": 0.6191188389453962,
"learning_rate": 2.165840302902632e-06,
"loss": 0.1979525566101074,
"step": 1840
},
{
"epoch": 0.592,
"grad_norm": 0.613998551941137,
"learning_rate": 2.1381821249625383e-06,
"loss": 0.2030627727508545,
"step": 1850
},
{
"epoch": 0.5952,
"grad_norm": 0.6115410126221079,
"learning_rate": 2.1105691070932465e-06,
"loss": 0.1951197624206543,
"step": 1860
},
{
"epoch": 0.5984,
"grad_norm": 0.5666967026000811,
"learning_rate": 2.083004695796238e-06,
"loss": 0.1926891803741455,
"step": 1870
},
{
"epoch": 0.6016,
"grad_norm": 0.5564168831256036,
"learning_rate": 2.055492331506194e-06,
"loss": 0.20087857246398927,
"step": 1880
},
{
"epoch": 0.6048,
"grad_norm": 0.647003695530594,
"learning_rate": 2.0280354481615814e-06,
"loss": 0.1991624116897583,
"step": 1890
},
{
"epoch": 0.608,
"grad_norm": 0.6020348842840653,
"learning_rate": 2.000637472776049e-06,
"loss": 0.20029563903808595,
"step": 1900
},
{
"epoch": 0.6112,
"grad_norm": 0.593460784828495,
"learning_rate": 1.973301825010685e-06,
"loss": 0.19462828636169432,
"step": 1910
},
{
"epoch": 0.6144,
"grad_norm": 0.6796900420369784,
"learning_rate": 1.9460319167471934e-06,
"loss": 0.20009157657623292,
"step": 1920
},
{
"epoch": 0.6176,
"grad_norm": 0.5803908647953272,
"learning_rate": 1.9188311516620466e-06,
"loss": 0.19473812580108643,
"step": 1930
},
{
"epoch": 0.6208,
"grad_norm": 0.5919196787967083,
"learning_rate": 1.891702924801651e-06,
"loss": 0.20190510749816895,
"step": 1940
},
{
"epoch": 0.624,
"grad_norm": 0.6056764566097385,
"learning_rate": 1.864650622158604e-06,
"loss": 0.2063821792602539,
"step": 1950
},
{
"epoch": 0.6272,
"grad_norm": 0.5106064574990916,
"learning_rate": 1.8376776202490666e-06,
"loss": 0.20139360427856445,
"step": 1960
},
{
"epoch": 0.6304,
"grad_norm": 0.5816570517079882,
"learning_rate": 1.8107872856913293e-06,
"loss": 0.19568054676055907,
"step": 1970
},
{
"epoch": 0.6336,
"grad_norm": 0.6100308085295513,
"learning_rate": 1.7839829747856096e-06,
"loss": 0.19661173820495606,
"step": 1980
},
{
"epoch": 0.6368,
"grad_norm": 0.6256775545767371,
"learning_rate": 1.7572680330951359e-06,
"loss": 0.19576869010925294,
"step": 1990
},
{
"epoch": 0.64,
"grad_norm": 0.5979254874380191,
"learning_rate": 1.7306457950285747e-06,
"loss": 0.19802470207214357,
"step": 2000
},
{
"epoch": 0.6432,
"grad_norm": 0.6445065470953916,
"learning_rate": 1.704119583423848e-06,
"loss": 0.19182772636413575,
"step": 2010
},
{
"epoch": 0.6464,
"grad_norm": 0.5238518416749739,
"learning_rate": 1.677692709133396e-06,
"loss": 0.19971816539764403,
"step": 2020
},
{
"epoch": 0.6496,
"grad_norm": 0.5902086462380663,
"learning_rate": 1.6513684706109311e-06,
"loss": 0.20058016777038573,
"step": 2030
},
{
"epoch": 0.6528,
"grad_norm": 0.5301315426540266,
"learning_rate": 1.6251501534997529e-06,
"loss": 0.19816763401031495,
"step": 2040
},
{
"epoch": 0.656,
"grad_norm": 0.5702221922649561,
"learning_rate": 1.5990410302226405e-06,
"loss": 0.19167234897613525,
"step": 2050
},
{
"epoch": 0.6592,
"grad_norm": 0.5682142108318351,
"learning_rate": 1.5730443595734162e-06,
"loss": 0.19806729555130004,
"step": 2060
},
{
"epoch": 0.6624,
"grad_norm": 0.6268750721579749,
"learning_rate": 1.5471633863101982e-06,
"loss": 0.1990320086479187,
"step": 2070
},
{
"epoch": 0.6656,
"grad_norm": 0.6501758398050216,
"learning_rate": 1.521401340750407e-06,
"loss": 0.20063567161560059,
"step": 2080
},
{
"epoch": 0.6688,
"grad_norm": 0.5367071332530153,
"learning_rate": 1.495761438367577e-06,
"loss": 0.2000502109527588,
"step": 2090
},
{
"epoch": 0.672,
"grad_norm": 0.6644202151690211,
"learning_rate": 1.4702468793900187e-06,
"loss": 0.19811663627624512,
"step": 2100
},
{
"epoch": 0.6752,
"grad_norm": 0.6020454013039992,
"learning_rate": 1.444860848401384e-06,
"loss": 0.19873985052108764,
"step": 2110
},
{
"epoch": 0.6784,
"grad_norm": 0.5672766014696592,
"learning_rate": 1.4196065139431866e-06,
"loss": 0.19663108587265016,
"step": 2120
},
{
"epoch": 0.6816,
"grad_norm": 0.6668756559032718,
"learning_rate": 1.3944870281193178e-06,
"loss": 0.19677751064300536,
"step": 2130
},
{
"epoch": 0.6848,
"grad_norm": 0.6146850263092741,
"learning_rate": 1.3695055262026208e-06,
"loss": 0.20252432823181152,
"step": 2140
},
{
"epoch": 0.688,
"grad_norm": 0.6023134400750195,
"learning_rate": 1.3446651262435679e-06,
"loss": 0.19564807415008545,
"step": 2150
},
{
"epoch": 0.6912,
"grad_norm": 0.5973758444267007,
"learning_rate": 1.3199689286810746e-06,
"loss": 0.19767165184020996,
"step": 2160
},
{
"epoch": 0.6944,
"grad_norm": 0.604085220565822,
"learning_rate": 1.2954200159555294e-06,
"loss": 0.19245314598083496,
"step": 2170
},
{
"epoch": 0.6976,
"grad_norm": 0.5971658440027723,
"learning_rate": 1.2710214521240527e-06,
"loss": 0.19593756198883056,
"step": 2180
},
{
"epoch": 0.7008,
"grad_norm": 0.6712656742168871,
"learning_rate": 1.246776282478063e-06,
"loss": 0.19848381280899047,
"step": 2190
},
{
"epoch": 0.704,
"grad_norm": 0.5303502593262494,
"learning_rate": 1.222687533163181e-06,
"loss": 0.19739968776702882,
"step": 2200
},
{
"epoch": 0.7072,
"grad_norm": 0.6329890536946617,
"learning_rate": 1.1987582108015228e-06,
"loss": 0.19885218143463135,
"step": 2210
},
{
"epoch": 0.7104,
"grad_norm": 0.6175733280769058,
"learning_rate": 1.1749913021164255e-06,
"loss": 0.20003676414489746,
"step": 2220
},
{
"epoch": 0.7136,
"grad_norm": 0.6297338992517326,
"learning_rate": 1.1513897735596702e-06,
"loss": 0.19420522451400757,
"step": 2230
},
{
"epoch": 0.7168,
"grad_norm": 0.5570261846558745,
"learning_rate": 1.127956570941218e-06,
"loss": 0.19144604206085206,
"step": 2240
},
{
"epoch": 0.72,
"grad_norm": 0.7464999016757174,
"learning_rate": 1.104694619061533e-06,
"loss": 0.20028018951416016,
"step": 2250
},
{
"epoch": 0.7232,
"grad_norm": 0.5813509472785208,
"learning_rate": 1.0816068213465295e-06,
"loss": 0.2022254228591919,
"step": 2260
},
{
"epoch": 0.7264,
"grad_norm": 0.5788680063085246,
"learning_rate": 1.0586960594851762e-06,
"loss": 0.19734264612197877,
"step": 2270
},
{
"epoch": 0.7296,
"grad_norm": 0.6879904092074834,
"learning_rate": 1.0359651930698217e-06,
"loss": 0.19566457271575927,
"step": 2280
},
{
"epoch": 0.7328,
"grad_norm": 0.545714278159425,
"learning_rate": 1.0134170592392837e-06,
"loss": 0.19808268547058105,
"step": 2290
},
{
"epoch": 0.736,
"grad_norm": 0.6957466724150051,
"learning_rate": 9.910544723247204e-07,
"loss": 0.19703471660614014,
"step": 2300
},
{
"epoch": 0.7392,
"grad_norm": 0.5722555379171206,
"learning_rate": 9.688802234983706e-07,
"loss": 0.19638856649398803,
"step": 2310
},
{
"epoch": 0.7424,
"grad_norm": 0.6657445816108672,
"learning_rate": 9.468970804251742e-07,
"loss": 0.1994560480117798,
"step": 2320
},
{
"epoch": 0.7456,
"grad_norm": 0.6118638240003964,
"learning_rate": 9.251077869173244e-07,
"loss": 0.19247424602508545,
"step": 2330
},
{
"epoch": 0.7488,
"grad_norm": 0.618262759129052,
"learning_rate": 9.035150625918054e-07,
"loss": 0.19384448528289794,
"step": 2340
},
{
"epoch": 0.752,
"grad_norm": 0.5841167908088344,
"learning_rate": 8.821216025309395e-07,
"loss": 0.19670048952102662,
"step": 2350
},
{
"epoch": 0.7552,
"grad_norm": 0.6330443090953268,
"learning_rate": 8.609300769460055e-07,
"loss": 0.191538667678833,
"step": 2360
},
{
"epoch": 0.7584,
"grad_norm": 0.6922248169402944,
"learning_rate": 8.399431308439592e-07,
"loss": 0.19869886636734008,
"step": 2370
},
{
"epoch": 0.7616,
"grad_norm": 0.5821907331028691,
"learning_rate": 8.191633836972962e-07,
"loss": 0.19837281703948975,
"step": 2380
},
{
"epoch": 0.7648,
"grad_norm": 0.5484553164447705,
"learning_rate": 7.985934291171024e-07,
"loss": 0.19366707801818847,
"step": 2390
},
{
"epoch": 0.768,
"grad_norm": 0.6131324978552078,
"learning_rate": 7.7823583452934e-07,
"loss": 0.19763607978820802,
"step": 2400
},
{
"epoch": 0.7712,
"grad_norm": 0.5665386766642198,
"learning_rate": 7.58093140854389e-07,
"loss": 0.19747262001037597,
"step": 2410
},
{
"epoch": 0.7744,
"grad_norm": 0.6702088035794936,
"learning_rate": 7.381678621899077e-07,
"loss": 0.19848825931549072,
"step": 2420
},
{
"epoch": 0.7776,
"grad_norm": 0.6808200224599221,
"learning_rate": 7.184624854970379e-07,
"loss": 0.19454023838043213,
"step": 2430
},
{
"epoch": 0.7808,
"grad_norm": 0.5446840545845119,
"learning_rate": 6.989794702899932e-07,
"loss": 0.1943270444869995,
"step": 2440
},
{
"epoch": 0.784,
"grad_norm": 0.6415010178339859,
"learning_rate": 6.797212483290777e-07,
"loss": 0.19584910869598388,
"step": 2450
},
{
"epoch": 0.7872,
"grad_norm": 0.603526871568268,
"learning_rate": 6.60690223317171e-07,
"loss": 0.19342836141586303,
"step": 2460
},
{
"epoch": 0.7904,
"grad_norm": 0.5817111419419255,
"learning_rate": 6.418887705997046e-07,
"loss": 0.19574793577194213,
"step": 2470
},
{
"epoch": 0.7936,
"grad_norm": 0.7792382444355755,
"learning_rate": 6.23319236868189e-07,
"loss": 0.1987607717514038,
"step": 2480
},
{
"epoch": 0.7968,
"grad_norm": 0.6291788716222239,
"learning_rate": 6.049839398673141e-07,
"loss": 0.20009655952453614,
"step": 2490
},
{
"epoch": 0.8,
"grad_norm": 0.674170182636883,
"learning_rate": 5.868851681056567e-07,
"loss": 0.2016763210296631,
"step": 2500
},
{
"epoch": 0.8032,
"grad_norm": 0.5738700746068163,
"learning_rate": 5.690251805700467e-07,
"loss": 0.19858623743057252,
"step": 2510
},
{
"epoch": 0.8064,
"grad_norm": 0.5748267344102337,
"learning_rate": 5.514062064436096e-07,
"loss": 0.19959205389022827,
"step": 2520
},
{
"epoch": 0.8096,
"grad_norm": 0.6464282974919533,
"learning_rate": 5.34030444827533e-07,
"loss": 0.19621236324310304,
"step": 2530
},
{
"epoch": 0.8128,
"grad_norm": 0.6390320405050175,
"learning_rate": 5.169000644665895e-07,
"loss": 0.19293551445007323,
"step": 2540
},
{
"epoch": 0.816,
"grad_norm": 0.5856228193289068,
"learning_rate": 5.000172034784442e-07,
"loss": 0.1952167272567749,
"step": 2550
},
{
"epoch": 0.8192,
"grad_norm": 0.6152721851543074,
"learning_rate": 4.833839690867853e-07,
"loss": 0.19755464792251587,
"step": 2560
},
{
"epoch": 0.8224,
"grad_norm": 0.6792777707129383,
"learning_rate": 4.6700243735831705e-07,
"loss": 0.1906466007232666,
"step": 2570
},
{
"epoch": 0.8256,
"grad_norm": 0.5650779115466599,
"learning_rate": 4.508746529436311e-07,
"loss": 0.1896218776702881,
"step": 2580
},
{
"epoch": 0.8288,
"grad_norm": 0.6068556104605155,
"learning_rate": 4.350026288220083e-07,
"loss": 0.1972370147705078,
"step": 2590
},
{
"epoch": 0.832,
"grad_norm": 0.6087844635927864,
"learning_rate": 4.1938834605017133e-07,
"loss": 0.19401493072509765,
"step": 2600
},
{
"epoch": 0.8352,
"grad_norm": 0.594443863161453,
"learning_rate": 4.0403375351501515e-07,
"loss": 0.19397275447845458,
"step": 2610
},
{
"epoch": 0.8384,
"grad_norm": 0.5777613928889838,
"learning_rate": 3.88940767690362e-07,
"loss": 0.19363962411880492,
"step": 2620
},
{
"epoch": 0.8416,
"grad_norm": 0.6122408540819826,
"learning_rate": 3.7411127239775774e-07,
"loss": 0.19224631786346436,
"step": 2630
},
{
"epoch": 0.8448,
"grad_norm": 0.5922115547592817,
"learning_rate": 3.595471185713431e-07,
"loss": 0.19027912616729736,
"step": 2640
},
{
"epoch": 0.848,
"grad_norm": 0.6012010067551694,
"learning_rate": 3.4525012402682826e-07,
"loss": 0.1921192765235901,
"step": 2650
},
{
"epoch": 0.8512,
"grad_norm": 0.6089446682050474,
"learning_rate": 3.3122207323460804e-07,
"loss": 0.19460537433624267,
"step": 2660
},
{
"epoch": 0.8544,
"grad_norm": 0.6314431181993275,
"learning_rate": 3.1746471709702963e-07,
"loss": 0.19075865745544435,
"step": 2670
},
{
"epoch": 0.8576,
"grad_norm": 0.6136529603147252,
"learning_rate": 3.039797727298585e-07,
"loss": 0.1973212718963623,
"step": 2680
},
{
"epoch": 0.8608,
"grad_norm": 0.6278068265217286,
"learning_rate": 2.9076892324795546e-07,
"loss": 0.19564627408981322,
"step": 2690
},
{
"epoch": 0.864,
"grad_norm": 0.6308491327804164,
"learning_rate": 2.778338175551995e-07,
"loss": 0.19089040756225586,
"step": 2700
},
{
"epoch": 0.8672,
"grad_norm": 0.6806226474068601,
"learning_rate": 2.6517607013868326e-07,
"loss": 0.19906394481658934,
"step": 2710
},
{
"epoch": 0.8704,
"grad_norm": 0.6497216896329614,
"learning_rate": 2.527972608672002e-07,
"loss": 0.19420729875564574,
"step": 2720
},
{
"epoch": 0.8736,
"grad_norm": 0.5988037888796804,
"learning_rate": 2.40698934794053e-07,
"loss": 0.1949334740638733,
"step": 2730
},
{
"epoch": 0.8768,
"grad_norm": 0.5825410688543936,
"learning_rate": 2.2888260196421237e-07,
"loss": 0.19373006820678712,
"step": 2740
},
{
"epoch": 0.88,
"grad_norm": 0.5659393573725252,
"learning_rate": 2.1734973722583735e-07,
"loss": 0.19743962287902833,
"step": 2750
},
{
"epoch": 0.8832,
"grad_norm": 0.6810045862821603,
"learning_rate": 2.0610178004619564e-07,
"loss": 0.18792747259140014,
"step": 2760
},
{
"epoch": 0.8864,
"grad_norm": 0.5624807528399969,
"learning_rate": 1.9514013433199834e-07,
"loss": 0.20065484046936036,
"step": 2770
},
{
"epoch": 0.8896,
"grad_norm": 0.5300049949985157,
"learning_rate": 1.8446616825416958e-07,
"loss": 0.19963890314102173,
"step": 2780
},
{
"epoch": 0.8928,
"grad_norm": 0.6417643354263414,
"learning_rate": 1.7408121407708007e-07,
"loss": 0.19946534633636476,
"step": 2790
},
{
"epoch": 0.896,
"grad_norm": 0.6263783317633913,
"learning_rate": 1.6398656799226253e-07,
"loss": 0.1873138427734375,
"step": 2800
},
{
"epoch": 0.8992,
"grad_norm": 0.6642472444356609,
"learning_rate": 1.5418348995662773e-07,
"loss": 0.1936098575592041,
"step": 2810
},
{
"epoch": 0.9024,
"grad_norm": 0.6361104958877116,
"learning_rate": 1.4467320353520275e-07,
"loss": 0.192909574508667,
"step": 2820
},
{
"epoch": 0.9056,
"grad_norm": 0.606401356191172,
"learning_rate": 1.3545689574841341e-07,
"loss": 0.1932598114013672,
"step": 2830
},
{
"epoch": 0.9088,
"grad_norm": 0.6138805257535019,
"learning_rate": 1.26535716923927e-07,
"loss": 0.19897468090057374,
"step": 2840
},
{
"epoch": 0.912,
"grad_norm": 0.6113791423952993,
"learning_rate": 1.1791078055307493e-07,
"loss": 0.19516528844833375,
"step": 2850
},
{
"epoch": 0.9152,
"grad_norm": 0.5897316619244026,
"learning_rate": 1.0958316315187289e-07,
"loss": 0.1947079300880432,
"step": 2860
},
{
"epoch": 0.9184,
"grad_norm": 0.6570448249633108,
"learning_rate": 1.0155390412665528e-07,
"loss": 0.19286593198776245,
"step": 2870
},
{
"epoch": 0.9216,
"grad_norm": 0.6143543897264965,
"learning_rate": 9.38240056443443e-08,
"loss": 0.18985612392425538,
"step": 2880
},
{
"epoch": 0.9248,
"grad_norm": 0.6208574768565508,
"learning_rate": 8.639443250736402e-08,
"loss": 0.1930636167526245,
"step": 2890
},
{
"epoch": 0.928,
"grad_norm": 0.6380337536968056,
"learning_rate": 7.926611203321777e-08,
"loss": 0.1940324306488037,
"step": 2900
},
{
"epoch": 0.9312,
"grad_norm": 0.6333119199427104,
"learning_rate": 7.243993393874882e-08,
"loss": 0.195207679271698,
"step": 2910
},
{
"epoch": 0.9344,
"grad_norm": 0.5601684784399228,
"learning_rate": 6.591675022908805e-08,
"loss": 0.1926344394683838,
"step": 2920
},
{
"epoch": 0.9376,
"grad_norm": 0.7001254632467586,
"learning_rate": 5.969737509131241e-08,
"loss": 0.189910888671875,
"step": 2930
},
{
"epoch": 0.9408,
"grad_norm": 0.5707165379372983,
"learning_rate": 5.3782584792823334e-08,
"loss": 0.1941395878791809,
"step": 2940
},
{
"epoch": 0.944,
"grad_norm": 0.637882100753534,
"learning_rate": 4.817311758445686e-08,
"loss": 0.19586544036865233,
"step": 2950
},
{
"epoch": 0.9472,
"grad_norm": 0.58305847153215,
"learning_rate": 4.286967360833866e-08,
"loss": 0.19621498584747316,
"step": 2960
},
{
"epoch": 0.9504,
"grad_norm": 0.6444124781946634,
"learning_rate": 3.787291481049754e-08,
"loss": 0.19597216844558715,
"step": 2970
},
{
"epoch": 0.9536,
"grad_norm": 0.68778482150424,
"learning_rate": 3.3183464858244364e-08,
"loss": 0.20229551792144776,
"step": 2980
},
{
"epoch": 0.9568,
"grad_norm": 0.589065287919965,
"learning_rate": 2.8801909062328992e-08,
"loss": 0.1879359722137451,
"step": 2990
},
{
"epoch": 0.96,
"grad_norm": 0.7200708770444023,
"learning_rate": 2.4728794303886248e-08,
"loss": 0.18806444406509398,
"step": 3000
},
{
"epoch": 0.9632,
"grad_norm": 0.6369212243333968,
"learning_rate": 2.0964628966175794e-08,
"loss": 0.19293060302734374,
"step": 3010
},
{
"epoch": 0.9664,
"grad_norm": 0.6150129328436796,
"learning_rate": 1.750988287113009e-08,
"loss": 0.19189660549163817,
"step": 3020
},
{
"epoch": 0.9696,
"grad_norm": 0.5966036549992078,
"learning_rate": 1.4364987220713278e-08,
"loss": 0.1992994427680969,
"step": 3030
},
{
"epoch": 0.9728,
"grad_norm": 0.6785615385564472,
"learning_rate": 1.1530334543099763e-08,
"loss": 0.19624128341674804,
"step": 3040
},
{
"epoch": 0.976,
"grad_norm": 0.626236262460755,
"learning_rate": 9.006278643683697e-09,
"loss": 0.19942662715911866,
"step": 3050
},
{
"epoch": 0.9792,
"grad_norm": 0.71228117768398,
"learning_rate": 6.793134560916514e-09,
"loss": 0.2007957935333252,
"step": 3060
},
{
"epoch": 0.9824,
"grad_norm": 0.5740813965788273,
"learning_rate": 4.891178526986451e-09,
"loss": 0.19730459451675414,
"step": 3070
},
{
"epoch": 0.9856,
"grad_norm": 0.6522249776214731,
"learning_rate": 3.3006479333413943e-09,
"loss": 0.1995969295501709,
"step": 3080
},
{
"epoch": 0.9888,
"grad_norm": 0.6484316026206892,
"learning_rate": 2.021741301058422e-09,
"loss": 0.19556543827056885,
"step": 3090
},
{
"epoch": 0.992,
"grad_norm": 0.6355663767406068,
"learning_rate": 1.0546182560652872e-09,
"loss": 0.19732578992843627,
"step": 3100
},
{
"epoch": 0.9952,
"grad_norm": 0.6169267731488666,
"learning_rate": 3.9939950921774607e-10,
"loss": 0.1917206883430481,
"step": 3110
},
{
"epoch": 0.9984,
"grad_norm": 0.5994063111681457,
"learning_rate": 5.616684123160854e-11,
"loss": 0.1916499137878418,
"step": 3120
},
{
"epoch": 1.0,
"step": 3125,
"total_flos": 1201860236279808.0,
"train_loss": 0.2128027264213562,
"train_runtime": 15463.9635,
"train_samples_per_second": 12.933,
"train_steps_per_second": 0.202
}
],
"logging_steps": 10,
"max_steps": 3125,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"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": 1201860236279808.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}