Safetensors
CosFly-Track / checkpoints /Qwen3.5-0.8B-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": 24.154499053955078,
"learning_rate": 1.437699680511182e-07,
"loss": 0.43009061813354493,
"step": 10
},
{
"epoch": 0.0064,
"grad_norm": 13.298259735107422,
"learning_rate": 3.0351437699680514e-07,
"loss": 0.4069969654083252,
"step": 20
},
{
"epoch": 0.0096,
"grad_norm": 4.6046857833862305,
"learning_rate": 4.6325878594249205e-07,
"loss": 0.3430281639099121,
"step": 30
},
{
"epoch": 0.0128,
"grad_norm": 2.4529078006744385,
"learning_rate": 6.230031948881789e-07,
"loss": 0.3079708099365234,
"step": 40
},
{
"epoch": 0.016,
"grad_norm": 1.7886693477630615,
"learning_rate": 7.82747603833866e-07,
"loss": 0.2815701484680176,
"step": 50
},
{
"epoch": 0.0192,
"grad_norm": 1.699796438217163,
"learning_rate": 9.424920127795528e-07,
"loss": 0.2694547653198242,
"step": 60
},
{
"epoch": 0.0224,
"grad_norm": 1.3927942514419556,
"learning_rate": 1.1022364217252397e-06,
"loss": 0.27406697273254393,
"step": 70
},
{
"epoch": 0.0256,
"grad_norm": 1.410854458808899,
"learning_rate": 1.2619808306709266e-06,
"loss": 0.25742459297180176,
"step": 80
},
{
"epoch": 0.0288,
"grad_norm": 1.304625391960144,
"learning_rate": 1.4217252396166134e-06,
"loss": 0.2527207851409912,
"step": 90
},
{
"epoch": 0.032,
"grad_norm": 1.3793327808380127,
"learning_rate": 1.5814696485623005e-06,
"loss": 0.2565553426742554,
"step": 100
},
{
"epoch": 0.0352,
"grad_norm": 1.3987141847610474,
"learning_rate": 1.7412140575079875e-06,
"loss": 0.25747051239013674,
"step": 110
},
{
"epoch": 0.0384,
"grad_norm": 1.3587315082550049,
"learning_rate": 1.9009584664536742e-06,
"loss": 0.24701259136199952,
"step": 120
},
{
"epoch": 0.0416,
"grad_norm": 1.4011516571044922,
"learning_rate": 2.060702875399361e-06,
"loss": 0.24983885288238525,
"step": 130
},
{
"epoch": 0.0448,
"grad_norm": 1.399139165878296,
"learning_rate": 2.220447284345048e-06,
"loss": 0.2414431095123291,
"step": 140
},
{
"epoch": 0.048,
"grad_norm": 1.3931989669799805,
"learning_rate": 2.380191693290735e-06,
"loss": 0.239495849609375,
"step": 150
},
{
"epoch": 0.0512,
"grad_norm": 1.2638698816299438,
"learning_rate": 2.539936102236422e-06,
"loss": 0.23560638427734376,
"step": 160
},
{
"epoch": 0.0544,
"grad_norm": 1.3124679327011108,
"learning_rate": 2.699680511182109e-06,
"loss": 0.2441340923309326,
"step": 170
},
{
"epoch": 0.0576,
"grad_norm": 1.3111082315444946,
"learning_rate": 2.8594249201277955e-06,
"loss": 0.2364725112915039,
"step": 180
},
{
"epoch": 0.0608,
"grad_norm": 1.3086109161376953,
"learning_rate": 3.0191693290734825e-06,
"loss": 0.24139461517333985,
"step": 190
},
{
"epoch": 0.064,
"grad_norm": 1.344279170036316,
"learning_rate": 3.17891373801917e-06,
"loss": 0.24218306541442872,
"step": 200
},
{
"epoch": 0.0672,
"grad_norm": 1.2795510292053223,
"learning_rate": 3.3386581469648564e-06,
"loss": 0.23907582759857177,
"step": 210
},
{
"epoch": 0.0704,
"grad_norm": 1.0937620401382446,
"learning_rate": 3.4984025559105434e-06,
"loss": 0.2335747480392456,
"step": 220
},
{
"epoch": 0.0736,
"grad_norm": 1.3082506656646729,
"learning_rate": 3.6581469648562303e-06,
"loss": 0.2359257698059082,
"step": 230
},
{
"epoch": 0.0768,
"grad_norm": 1.1195415258407593,
"learning_rate": 3.817891373801918e-06,
"loss": 0.23113086223602294,
"step": 240
},
{
"epoch": 0.08,
"grad_norm": 1.0815153121948242,
"learning_rate": 3.977635782747604e-06,
"loss": 0.22873566150665284,
"step": 250
},
{
"epoch": 0.0832,
"grad_norm": 1.2189197540283203,
"learning_rate": 4.137380191693291e-06,
"loss": 0.2301633596420288,
"step": 260
},
{
"epoch": 0.0864,
"grad_norm": 1.0007416009902954,
"learning_rate": 4.297124600638978e-06,
"loss": 0.23275952339172362,
"step": 270
},
{
"epoch": 0.0896,
"grad_norm": 1.0670294761657715,
"learning_rate": 4.456869009584665e-06,
"loss": 0.22844791412353516,
"step": 280
},
{
"epoch": 0.0928,
"grad_norm": 1.0621100664138794,
"learning_rate": 4.616613418530352e-06,
"loss": 0.23461694717407228,
"step": 290
},
{
"epoch": 0.096,
"grad_norm": 1.184499740600586,
"learning_rate": 4.776357827476039e-06,
"loss": 0.23582732677459717,
"step": 300
},
{
"epoch": 0.0992,
"grad_norm": 1.1273094415664673,
"learning_rate": 4.936102236421725e-06,
"loss": 0.22369306087493895,
"step": 310
},
{
"epoch": 0.1024,
"grad_norm": 0.982417106628418,
"learning_rate": 4.999943833158769e-06,
"loss": 0.2245041847229004,
"step": 320
},
{
"epoch": 0.1056,
"grad_norm": 1.0180294513702393,
"learning_rate": 4.999600600490783e-06,
"loss": 0.2303633213043213,
"step": 330
},
{
"epoch": 0.1088,
"grad_norm": 1.0338643789291382,
"learning_rate": 4.9989453817439345e-06,
"loss": 0.22460854053497314,
"step": 340
},
{
"epoch": 0.112,
"grad_norm": 0.8967295289039612,
"learning_rate": 4.997978258698942e-06,
"loss": 0.224709415435791,
"step": 350
},
{
"epoch": 0.1152,
"grad_norm": 1.074678897857666,
"learning_rate": 4.996699352066659e-06,
"loss": 0.22632455825805664,
"step": 360
},
{
"epoch": 0.1184,
"grad_norm": 0.9827248454093933,
"learning_rate": 4.995108821473014e-06,
"loss": 0.22602491378784179,
"step": 370
},
{
"epoch": 0.1216,
"grad_norm": 0.954534649848938,
"learning_rate": 4.993206865439084e-06,
"loss": 0.2213623046875,
"step": 380
},
{
"epoch": 0.1248,
"grad_norm": 0.8947958946228027,
"learning_rate": 4.990993721356317e-06,
"loss": 0.22030019760131836,
"step": 390
},
{
"epoch": 0.128,
"grad_norm": 0.994922399520874,
"learning_rate": 4.988469665456901e-06,
"loss": 0.22146718502044677,
"step": 400
},
{
"epoch": 0.1312,
"grad_norm": 0.960228443145752,
"learning_rate": 4.985635012779288e-06,
"loss": 0.2290752649307251,
"step": 410
},
{
"epoch": 0.1344,
"grad_norm": 0.8288887739181519,
"learning_rate": 4.98249011712887e-06,
"loss": 0.22198934555053712,
"step": 420
},
{
"epoch": 0.1376,
"grad_norm": 0.9185967445373535,
"learning_rate": 4.979035371033824e-06,
"loss": 0.22025692462921143,
"step": 430
},
{
"epoch": 0.1408,
"grad_norm": 0.9180241227149963,
"learning_rate": 4.975271205696115e-06,
"loss": 0.22060942649841309,
"step": 440
},
{
"epoch": 0.144,
"grad_norm": 0.9974319338798523,
"learning_rate": 4.971198090937671e-06,
"loss": 0.21446082592010499,
"step": 450
},
{
"epoch": 0.1472,
"grad_norm": 1.0199272632598877,
"learning_rate": 4.966816535141756e-06,
"loss": 0.21097931861877442,
"step": 460
},
{
"epoch": 0.1504,
"grad_norm": 1.0272812843322754,
"learning_rate": 4.9621270851895035e-06,
"loss": 0.22108638286590576,
"step": 470
},
{
"epoch": 0.1536,
"grad_norm": 0.969963014125824,
"learning_rate": 4.957130326391662e-06,
"loss": 0.22175126075744628,
"step": 480
},
{
"epoch": 0.1568,
"grad_norm": 0.981428325176239,
"learning_rate": 4.951826882415544e-06,
"loss": 0.21852502822875977,
"step": 490
},
{
"epoch": 0.16,
"grad_norm": 1.0144567489624023,
"learning_rate": 4.946217415207177e-06,
"loss": 0.2126768112182617,
"step": 500
},
{
"epoch": 0.1632,
"grad_norm": 0.8966764807701111,
"learning_rate": 4.940302624908689e-06,
"loss": 0.21713790893554688,
"step": 510
},
{
"epoch": 0.1664,
"grad_norm": 1.0284103155136108,
"learning_rate": 4.934083249770912e-06,
"loss": 0.21167585849761963,
"step": 520
},
{
"epoch": 0.1696,
"grad_norm": 0.9075572490692139,
"learning_rate": 4.927560066061251e-06,
"loss": 0.21518006324768066,
"step": 530
},
{
"epoch": 0.1728,
"grad_norm": 0.8963767886161804,
"learning_rate": 4.920733887966783e-06,
"loss": 0.22648308277130128,
"step": 540
},
{
"epoch": 0.176,
"grad_norm": 0.8435696363449097,
"learning_rate": 4.913605567492636e-06,
"loss": 0.2141176223754883,
"step": 550
},
{
"epoch": 0.1792,
"grad_norm": 0.864231526851654,
"learning_rate": 4.906175994355656e-06,
"loss": 0.21634316444396973,
"step": 560
},
{
"epoch": 0.1824,
"grad_norm": 1.0810909271240234,
"learning_rate": 4.898446095873345e-06,
"loss": 0.2190020799636841,
"step": 570
},
{
"epoch": 0.1856,
"grad_norm": 0.9246620535850525,
"learning_rate": 4.890416836848128e-06,
"loss": 0.21658658981323242,
"step": 580
},
{
"epoch": 0.1888,
"grad_norm": 0.9349477291107178,
"learning_rate": 4.882089219446925e-06,
"loss": 0.21075100898742677,
"step": 590
},
{
"epoch": 0.192,
"grad_norm": 1.0043901205062866,
"learning_rate": 4.873464283076074e-06,
"loss": 0.21601030826568604,
"step": 600
},
{
"epoch": 0.1952,
"grad_norm": 0.8367205858230591,
"learning_rate": 4.864543104251587e-06,
"loss": 0.2150674819946289,
"step": 610
},
{
"epoch": 0.1984,
"grad_norm": 0.9035486578941345,
"learning_rate": 4.855326796464798e-06,
"loss": 0.21784515380859376,
"step": 620
},
{
"epoch": 0.2016,
"grad_norm": 0.8648081421852112,
"learning_rate": 4.8458165100433725e-06,
"loss": 0.21317291259765625,
"step": 630
},
{
"epoch": 0.2048,
"grad_norm": 0.9622132182121277,
"learning_rate": 4.836013432007738e-06,
"loss": 0.21332902908325196,
"step": 640
},
{
"epoch": 0.208,
"grad_norm": 0.8541603684425354,
"learning_rate": 4.825918785922921e-06,
"loss": 0.21687707901000977,
"step": 650
},
{
"epoch": 0.2112,
"grad_norm": 0.9410703182220459,
"learning_rate": 4.8155338317458315e-06,
"loss": 0.21425023078918456,
"step": 660
},
{
"epoch": 0.2144,
"grad_norm": 0.8768807649612427,
"learning_rate": 4.804859865668002e-06,
"loss": 0.21114716529846192,
"step": 670
},
{
"epoch": 0.2176,
"grad_norm": 0.9346839785575867,
"learning_rate": 4.793898219953804e-06,
"loss": 0.21280355453491212,
"step": 680
},
{
"epoch": 0.2208,
"grad_norm": 0.773104727268219,
"learning_rate": 4.782650262774164e-06,
"loss": 0.21328551769256593,
"step": 690
},
{
"epoch": 0.224,
"grad_norm": 0.8742129802703857,
"learning_rate": 4.7711173980357886e-06,
"loss": 0.2097193717956543,
"step": 700
},
{
"epoch": 0.2272,
"grad_norm": 0.851413905620575,
"learning_rate": 4.759301065205947e-06,
"loss": 0.21059794425964357,
"step": 710
},
{
"epoch": 0.2304,
"grad_norm": 0.9423609972000122,
"learning_rate": 4.7472027391328e-06,
"loss": 0.2117161273956299,
"step": 720
},
{
"epoch": 0.2336,
"grad_norm": 1.0031023025512695,
"learning_rate": 4.734823929861317e-06,
"loss": 0.21474902629852294,
"step": 730
},
{
"epoch": 0.2368,
"grad_norm": 0.9702557325363159,
"learning_rate": 4.722166182444801e-06,
"loss": 0.2112131357192993,
"step": 740
},
{
"epoch": 0.24,
"grad_norm": 0.8935959935188293,
"learning_rate": 4.709231076752045e-06,
"loss": 0.2096531867980957,
"step": 750
},
{
"epoch": 0.2432,
"grad_norm": 0.8280373215675354,
"learning_rate": 4.696020227270142e-06,
"loss": 0.2149428367614746,
"step": 760
},
{
"epoch": 0.2464,
"grad_norm": 0.8387318849563599,
"learning_rate": 4.6825352829029705e-06,
"loss": 0.21069800853729248,
"step": 770
},
{
"epoch": 0.2496,
"grad_norm": 0.9887408018112183,
"learning_rate": 4.668777926765392e-06,
"loss": 0.20976176261901855,
"step": 780
},
{
"epoch": 0.2528,
"grad_norm": 0.9188724756240845,
"learning_rate": 4.6547498759731725e-06,
"loss": 0.20372920036315917,
"step": 790
},
{
"epoch": 0.256,
"grad_norm": 0.8613991737365723,
"learning_rate": 4.6404528814286575e-06,
"loss": 0.20750484466552735,
"step": 800
},
{
"epoch": 0.2592,
"grad_norm": 0.8634099960327148,
"learning_rate": 4.6258887276022425e-06,
"loss": 0.21537406444549562,
"step": 810
},
{
"epoch": 0.2624,
"grad_norm": 1.0032637119293213,
"learning_rate": 4.611059232309639e-06,
"loss": 0.20938510894775392,
"step": 820
},
{
"epoch": 0.2656,
"grad_norm": 1.013604998588562,
"learning_rate": 4.595966246484986e-06,
"loss": 0.21059298515319824,
"step": 830
},
{
"epoch": 0.2688,
"grad_norm": 0.884544312953949,
"learning_rate": 4.580611653949829e-06,
"loss": 0.2108741283416748,
"step": 840
},
{
"epoch": 0.272,
"grad_norm": 0.8698775768280029,
"learning_rate": 4.564997371177992e-06,
"loss": 0.2090519905090332,
"step": 850
},
{
"epoch": 0.2752,
"grad_norm": 0.883059024810791,
"learning_rate": 4.54912534705637e-06,
"loss": 0.20870964527130126,
"step": 860
},
{
"epoch": 0.2784,
"grad_norm": 0.8841277360916138,
"learning_rate": 4.532997562641683e-06,
"loss": 0.20556044578552246,
"step": 870
},
{
"epoch": 0.2816,
"grad_norm": 0.9181431531906128,
"learning_rate": 4.516616030913214e-06,
"loss": 0.20940375328063965,
"step": 880
},
{
"epoch": 0.2848,
"grad_norm": 0.8961875438690186,
"learning_rate": 4.499982796521556e-06,
"loss": 0.20449237823486327,
"step": 890
},
{
"epoch": 0.288,
"grad_norm": 0.9166404604911804,
"learning_rate": 4.48309993553341e-06,
"loss": 0.20691614151000975,
"step": 900
},
{
"epoch": 0.2912,
"grad_norm": 0.8640049695968628,
"learning_rate": 4.465969555172468e-06,
"loss": 0.20648136138916015,
"step": 910
},
{
"epoch": 0.2944,
"grad_norm": 0.8916935324668884,
"learning_rate": 4.448593793556391e-06,
"loss": 0.2128384828567505,
"step": 920
},
{
"epoch": 0.2976,
"grad_norm": 0.9438337087631226,
"learning_rate": 4.430974819429954e-06,
"loss": 0.20601146221160888,
"step": 930
},
{
"epoch": 0.3008,
"grad_norm": 0.8253967761993408,
"learning_rate": 4.413114831894344e-06,
"loss": 0.20144739151000976,
"step": 940
},
{
"epoch": 0.304,
"grad_norm": 0.8320073485374451,
"learning_rate": 4.3950160601326865e-06,
"loss": 0.1991488218307495,
"step": 950
},
{
"epoch": 0.3072,
"grad_norm": 0.9561188817024231,
"learning_rate": 4.376680763131811e-06,
"loss": 0.20644969940185548,
"step": 960
},
{
"epoch": 0.3104,
"grad_norm": 0.803125262260437,
"learning_rate": 4.358111229400296e-06,
"loss": 0.20915827751159669,
"step": 970
},
{
"epoch": 0.3136,
"grad_norm": 0.8132813572883606,
"learning_rate": 4.33930977668283e-06,
"loss": 0.21236319541931153,
"step": 980
},
{
"epoch": 0.3168,
"grad_norm": 0.9199708700180054,
"learning_rate": 4.320278751670922e-06,
"loss": 0.2055152416229248,
"step": 990
},
{
"epoch": 0.32,
"grad_norm": 0.8928452134132385,
"learning_rate": 4.301020529710009e-06,
"loss": 0.206940221786499,
"step": 1000
},
{
"epoch": 0.3232,
"grad_norm": 0.9517531991004944,
"learning_rate": 4.281537514502962e-06,
"loss": 0.2108684539794922,
"step": 1010
},
{
"epoch": 0.3264,
"grad_norm": 0.9454931020736694,
"learning_rate": 4.261832137810093e-06,
"loss": 0.20762853622436522,
"step": 1020
},
{
"epoch": 0.3296,
"grad_norm": 0.8216907978057861,
"learning_rate": 4.241906859145611e-06,
"loss": 0.2080090045928955,
"step": 1030
},
{
"epoch": 0.3328,
"grad_norm": 0.7746848464012146,
"learning_rate": 4.221764165470661e-06,
"loss": 0.2045358896255493,
"step": 1040
},
{
"epoch": 0.336,
"grad_norm": 0.9690631031990051,
"learning_rate": 4.201406570882898e-06,
"loss": 0.2042619466781616,
"step": 1050
},
{
"epoch": 0.3392,
"grad_norm": 0.8224774599075317,
"learning_rate": 4.180836616302704e-06,
"loss": 0.20257024765014647,
"step": 1060
},
{
"epoch": 0.3424,
"grad_norm": 1.0545451641082764,
"learning_rate": 4.160056869156041e-06,
"loss": 0.20705747604370117,
"step": 1070
},
{
"epoch": 0.3456,
"grad_norm": 0.7687672972679138,
"learning_rate": 4.139069923053995e-06,
"loss": 0.20620079040527345,
"step": 1080
},
{
"epoch": 0.3488,
"grad_norm": 0.8217872977256775,
"learning_rate": 4.117878397469062e-06,
"loss": 0.2085184097290039,
"step": 1090
},
{
"epoch": 0.352,
"grad_norm": 0.8416702747344971,
"learning_rate": 4.096484937408195e-06,
"loss": 0.2004465103149414,
"step": 1100
},
{
"epoch": 0.3552,
"grad_norm": 0.9369352459907532,
"learning_rate": 4.074892213082676e-06,
"loss": 0.2000356674194336,
"step": 1110
},
{
"epoch": 0.3584,
"grad_norm": 0.8932799100875854,
"learning_rate": 4.0531029195748265e-06,
"loss": 0.2081597328186035,
"step": 1120
},
{
"epoch": 0.3616,
"grad_norm": 0.8241865634918213,
"learning_rate": 4.03111977650163e-06,
"loss": 0.20646026134490966,
"step": 1130
},
{
"epoch": 0.3648,
"grad_norm": 0.895818829536438,
"learning_rate": 4.008945527675281e-06,
"loss": 0.20636439323425293,
"step": 1140
},
{
"epoch": 0.368,
"grad_norm": 1.025847315788269,
"learning_rate": 3.986582940760717e-06,
"loss": 0.19549331665039063,
"step": 1150
},
{
"epoch": 0.3712,
"grad_norm": 0.8418158292770386,
"learning_rate": 3.9640348069301785e-06,
"loss": 0.2030022621154785,
"step": 1160
},
{
"epoch": 0.3744,
"grad_norm": 0.8930888175964355,
"learning_rate": 3.941303940514826e-06,
"loss": 0.2045799732208252,
"step": 1170
},
{
"epoch": 0.3776,
"grad_norm": 0.9579597115516663,
"learning_rate": 3.918393178653472e-06,
"loss": 0.2056349515914917,
"step": 1180
},
{
"epoch": 0.3808,
"grad_norm": 0.8847565054893494,
"learning_rate": 3.895305380938468e-06,
"loss": 0.2011105537414551,
"step": 1190
},
{
"epoch": 0.384,
"grad_norm": 0.8585783839225769,
"learning_rate": 3.872043429058783e-06,
"loss": 0.199721622467041,
"step": 1200
},
{
"epoch": 0.3872,
"grad_norm": 0.8783968091011047,
"learning_rate": 3.84861022644033e-06,
"loss": 0.20214991569519042,
"step": 1210
},
{
"epoch": 0.3904,
"grad_norm": 0.814146101474762,
"learning_rate": 3.825008697883574e-06,
"loss": 0.21019458770751953,
"step": 1220
},
{
"epoch": 0.3936,
"grad_norm": 0.808813214302063,
"learning_rate": 3.8012417891984776e-06,
"loss": 0.20353050231933595,
"step": 1230
},
{
"epoch": 0.3968,
"grad_norm": 0.86122727394104,
"learning_rate": 3.777312466836819e-06,
"loss": 0.20200772285461427,
"step": 1240
},
{
"epoch": 0.4,
"grad_norm": 0.8682144284248352,
"learning_rate": 3.7532237175219378e-06,
"loss": 0.2004477024078369,
"step": 1250
},
{
"epoch": 0.4032,
"grad_norm": 0.9007877707481384,
"learning_rate": 3.728978547875948e-06,
"loss": 0.2053370952606201,
"step": 1260
},
{
"epoch": 0.4064,
"grad_norm": 0.8896200060844421,
"learning_rate": 3.7045799840444712e-06,
"loss": 0.19951150417327881,
"step": 1270
},
{
"epoch": 0.4096,
"grad_norm": 0.951781690120697,
"learning_rate": 3.6800310713189258e-06,
"loss": 0.20167291164398193,
"step": 1280
},
{
"epoch": 0.4128,
"grad_norm": 0.9838011860847473,
"learning_rate": 3.6553348737564328e-06,
"loss": 0.20190558433532715,
"step": 1290
},
{
"epoch": 0.416,
"grad_norm": 0.7393888235092163,
"learning_rate": 3.6304944737973794e-06,
"loss": 0.20698275566101074,
"step": 1300
},
{
"epoch": 0.4192,
"grad_norm": 0.9344687461853027,
"learning_rate": 3.6055129718806836e-06,
"loss": 0.2017432689666748,
"step": 1310
},
{
"epoch": 0.4224,
"grad_norm": 0.8322877287864685,
"learning_rate": 3.5803934860568134e-06,
"loss": 0.19672255516052245,
"step": 1320
},
{
"epoch": 0.4256,
"grad_norm": 0.8978738784790039,
"learning_rate": 3.5551391515986163e-06,
"loss": 0.20154342651367188,
"step": 1330
},
{
"epoch": 0.4288,
"grad_norm": 0.9294552206993103,
"learning_rate": 3.529753120609982e-06,
"loss": 0.19790271520614625,
"step": 1340
},
{
"epoch": 0.432,
"grad_norm": 0.9712645411491394,
"learning_rate": 3.5042385616324243e-06,
"loss": 0.2009965419769287,
"step": 1350
},
{
"epoch": 0.4352,
"grad_norm": 0.7751940488815308,
"learning_rate": 3.4785986592495934e-06,
"loss": 0.19958891868591308,
"step": 1360
},
{
"epoch": 0.4384,
"grad_norm": 0.7986553311347961,
"learning_rate": 3.452836613689803e-06,
"loss": 0.19687262773513795,
"step": 1370
},
{
"epoch": 0.4416,
"grad_norm": 0.7831050753593445,
"learning_rate": 3.426955640426584e-06,
"loss": 0.20044875144958496,
"step": 1380
},
{
"epoch": 0.4448,
"grad_norm": 0.8416183590888977,
"learning_rate": 3.4009589697773605e-06,
"loss": 0.1997106909751892,
"step": 1390
},
{
"epoch": 0.448,
"grad_norm": 0.9130992293357849,
"learning_rate": 3.3748498465002475e-06,
"loss": 0.19710896015167237,
"step": 1400
},
{
"epoch": 0.4512,
"grad_norm": 0.8053887486457825,
"learning_rate": 3.3486315293890693e-06,
"loss": 0.20499968528747559,
"step": 1410
},
{
"epoch": 0.4544,
"grad_norm": 0.8130092024803162,
"learning_rate": 3.3223072908666053e-06,
"loss": 0.1947783946990967,
"step": 1420
},
{
"epoch": 0.4576,
"grad_norm": 0.8554696440696716,
"learning_rate": 3.295880416576153e-06,
"loss": 0.20600056648254395,
"step": 1430
},
{
"epoch": 0.4608,
"grad_norm": 0.8399017453193665,
"learning_rate": 3.269354204971427e-06,
"loss": 0.19943264722824097,
"step": 1440
},
{
"epoch": 0.464,
"grad_norm": 0.9924997687339783,
"learning_rate": 3.242731966904865e-06,
"loss": 0.19647090435028075,
"step": 1450
},
{
"epoch": 0.4672,
"grad_norm": 0.8956066966056824,
"learning_rate": 3.2160170252143913e-06,
"loss": 0.19706339836120607,
"step": 1460
},
{
"epoch": 0.4704,
"grad_norm": 0.773949384689331,
"learning_rate": 3.1892127143086716e-06,
"loss": 0.20296840667724608,
"step": 1470
},
{
"epoch": 0.4736,
"grad_norm": 0.9293017983436584,
"learning_rate": 3.1623223797509347e-06,
"loss": 0.19287054538726806,
"step": 1480
},
{
"epoch": 0.4768,
"grad_norm": 1.0113096237182617,
"learning_rate": 3.135349377841396e-06,
"loss": 0.19713553190231323,
"step": 1490
},
{
"epoch": 0.48,
"grad_norm": 0.8387943506240845,
"learning_rate": 3.1082970751983497e-06,
"loss": 0.2030486583709717,
"step": 1500
},
{
"epoch": 0.4832,
"grad_norm": 0.8359475135803223,
"learning_rate": 3.0811688483379546e-06,
"loss": 0.20096054077148437,
"step": 1510
},
{
"epoch": 0.4864,
"grad_norm": 0.8501754999160767,
"learning_rate": 3.0539680832528074e-06,
"loss": 0.2003248691558838,
"step": 1520
},
{
"epoch": 0.4896,
"grad_norm": 0.8864139318466187,
"learning_rate": 3.026698174989316e-06,
"loss": 0.19884690046310424,
"step": 1530
},
{
"epoch": 0.4928,
"grad_norm": 0.8472388982772827,
"learning_rate": 2.999362527223952e-06,
"loss": 0.19895727634429933,
"step": 1540
},
{
"epoch": 0.496,
"grad_norm": 0.835716187953949,
"learning_rate": 2.9719645518384194e-06,
"loss": 0.20096054077148437,
"step": 1550
},
{
"epoch": 0.4992,
"grad_norm": 1.056063175201416,
"learning_rate": 2.944507668493807e-06,
"loss": 0.20421340465545654,
"step": 1560
},
{
"epoch": 0.5024,
"grad_norm": 0.8620187044143677,
"learning_rate": 2.9169953042037623e-06,
"loss": 0.1987598180770874,
"step": 1570
},
{
"epoch": 0.5056,
"grad_norm": 0.9530871510505676,
"learning_rate": 2.889430892906754e-06,
"loss": 0.19546263217926024,
"step": 1580
},
{
"epoch": 0.5088,
"grad_norm": 0.9080846309661865,
"learning_rate": 2.861817875037462e-06,
"loss": 0.1925678253173828,
"step": 1590
},
{
"epoch": 0.512,
"grad_norm": 0.8608678579330444,
"learning_rate": 2.8341596970973683e-06,
"loss": 0.2030557632446289,
"step": 1600
},
{
"epoch": 0.5152,
"grad_norm": 0.8551720380783081,
"learning_rate": 2.80645981122458e-06,
"loss": 0.1977646231651306,
"step": 1610
},
{
"epoch": 0.5184,
"grad_norm": 0.8102371692657471,
"learning_rate": 2.7787216747629508e-06,
"loss": 0.20460376739501954,
"step": 1620
},
{
"epoch": 0.5216,
"grad_norm": 0.757784366607666,
"learning_rate": 2.7509487498305615e-06,
"loss": 0.1916171431541443,
"step": 1630
},
{
"epoch": 0.5248,
"grad_norm": 0.8284843564033508,
"learning_rate": 2.7231445028875924e-06,
"loss": 0.19816431999206544,
"step": 1640
},
{
"epoch": 0.528,
"grad_norm": 0.8836312890052795,
"learning_rate": 2.6953124043036604e-06,
"loss": 0.19672504663467408,
"step": 1650
},
{
"epoch": 0.5312,
"grad_norm": 0.853327214717865,
"learning_rate": 2.667455927924667e-06,
"loss": 0.1962347984313965,
"step": 1660
},
{
"epoch": 0.5344,
"grad_norm": 0.7774537205696106,
"learning_rate": 2.6395785506392164e-06,
"loss": 0.19192854166030884,
"step": 1670
},
{
"epoch": 0.5376,
"grad_norm": 0.9308248162269592,
"learning_rate": 2.6116837519446407e-06,
"loss": 0.194757080078125,
"step": 1680
},
{
"epoch": 0.5408,
"grad_norm": 0.8841214776039124,
"learning_rate": 2.5837750135127192e-06,
"loss": 0.19211833477020263,
"step": 1690
},
{
"epoch": 0.544,
"grad_norm": 0.8462106585502625,
"learning_rate": 2.555855818755108e-06,
"loss": 0.19800645112991333,
"step": 1700
},
{
"epoch": 0.5472,
"grad_norm": 0.954951286315918,
"learning_rate": 2.5279296523885636e-06,
"loss": 0.19544832706451415,
"step": 1710
},
{
"epoch": 0.5504,
"grad_norm": 0.9142067432403564,
"learning_rate": 2.5e-06,
"loss": 0.19691340923309325,
"step": 1720
},
{
"epoch": 0.5536,
"grad_norm": 0.8898570537567139,
"learning_rate": 2.472070347611437e-06,
"loss": 0.1904543399810791,
"step": 1730
},
{
"epoch": 0.5568,
"grad_norm": 0.8316683769226074,
"learning_rate": 2.444144181244893e-06,
"loss": 0.19625117778778076,
"step": 1740
},
{
"epoch": 0.56,
"grad_norm": 0.8947731852531433,
"learning_rate": 2.416224986487282e-06,
"loss": 0.1924436092376709,
"step": 1750
},
{
"epoch": 0.5632,
"grad_norm": 0.8805821537971497,
"learning_rate": 2.3883162480553605e-06,
"loss": 0.19019029140472413,
"step": 1760
},
{
"epoch": 0.5664,
"grad_norm": 0.9083222150802612,
"learning_rate": 2.3604214493607844e-06,
"loss": 0.19431073665618898,
"step": 1770
},
{
"epoch": 0.5696,
"grad_norm": 0.9499437212944031,
"learning_rate": 2.332544072075333e-06,
"loss": 0.19770257472991942,
"step": 1780
},
{
"epoch": 0.5728,
"grad_norm": 0.8856005668640137,
"learning_rate": 2.30468759569634e-06,
"loss": 0.1973590612411499,
"step": 1790
},
{
"epoch": 0.576,
"grad_norm": 0.930976152420044,
"learning_rate": 2.276855497112408e-06,
"loss": 0.1932486414909363,
"step": 1800
},
{
"epoch": 0.5792,
"grad_norm": 0.8594582676887512,
"learning_rate": 2.2490512501694394e-06,
"loss": 0.18842577934265137,
"step": 1810
},
{
"epoch": 0.5824,
"grad_norm": 0.8443211913108826,
"learning_rate": 2.2212783252370496e-06,
"loss": 0.19471209049224852,
"step": 1820
},
{
"epoch": 0.5856,
"grad_norm": 0.8561191558837891,
"learning_rate": 2.1935401887754213e-06,
"loss": 0.1995155930519104,
"step": 1830
},
{
"epoch": 0.5888,
"grad_norm": 0.8672814965248108,
"learning_rate": 2.165840302902632e-06,
"loss": 0.19246792793273926,
"step": 1840
},
{
"epoch": 0.592,
"grad_norm": 0.9349257349967957,
"learning_rate": 2.1381821249625383e-06,
"loss": 0.1974799871444702,
"step": 1850
},
{
"epoch": 0.5952,
"grad_norm": 0.9783422946929932,
"learning_rate": 2.1105691070932465e-06,
"loss": 0.18990163803100585,
"step": 1860
},
{
"epoch": 0.5984,
"grad_norm": 0.9529528617858887,
"learning_rate": 2.083004695796238e-06,
"loss": 0.18725800514221191,
"step": 1870
},
{
"epoch": 0.6016,
"grad_norm": 0.9632482528686523,
"learning_rate": 2.055492331506194e-06,
"loss": 0.19491477012634278,
"step": 1880
},
{
"epoch": 0.6048,
"grad_norm": 0.9552566409111023,
"learning_rate": 2.0280354481615814e-06,
"loss": 0.19355373382568358,
"step": 1890
},
{
"epoch": 0.608,
"grad_norm": 0.8642057776451111,
"learning_rate": 2.000637472776049e-06,
"loss": 0.19464772939682007,
"step": 1900
},
{
"epoch": 0.6112,
"grad_norm": 0.9171792268753052,
"learning_rate": 1.973301825010685e-06,
"loss": 0.18813092708587648,
"step": 1910
},
{
"epoch": 0.6144,
"grad_norm": 0.9625651836395264,
"learning_rate": 1.9460319167471934e-06,
"loss": 0.19436421394348144,
"step": 1920
},
{
"epoch": 0.6176,
"grad_norm": 0.8342574238777161,
"learning_rate": 1.9188311516620466e-06,
"loss": 0.18922022581100464,
"step": 1930
},
{
"epoch": 0.6208,
"grad_norm": 0.9471722841262817,
"learning_rate": 1.891702924801651e-06,
"loss": 0.19615614414215088,
"step": 1940
},
{
"epoch": 0.624,
"grad_norm": 0.9361758232116699,
"learning_rate": 1.864650622158604e-06,
"loss": 0.20051980018615723,
"step": 1950
},
{
"epoch": 0.6272,
"grad_norm": 0.8602835536003113,
"learning_rate": 1.8376776202490666e-06,
"loss": 0.19571347236633302,
"step": 1960
},
{
"epoch": 0.6304,
"grad_norm": 0.9917039275169373,
"learning_rate": 1.8107872856913293e-06,
"loss": 0.19011812210083007,
"step": 1970
},
{
"epoch": 0.6336,
"grad_norm": 0.7912291884422302,
"learning_rate": 1.7839829747856096e-06,
"loss": 0.1915016531944275,
"step": 1980
},
{
"epoch": 0.6368,
"grad_norm": 0.9529620409011841,
"learning_rate": 1.7572680330951359e-06,
"loss": 0.19014396667480468,
"step": 1990
},
{
"epoch": 0.64,
"grad_norm": 0.8650413751602173,
"learning_rate": 1.7306457950285747e-06,
"loss": 0.19222151041030883,
"step": 2000
},
{
"epoch": 0.6432,
"grad_norm": 0.9507162570953369,
"learning_rate": 1.704119583423848e-06,
"loss": 0.1860235333442688,
"step": 2010
},
{
"epoch": 0.6464,
"grad_norm": 0.8666033744812012,
"learning_rate": 1.677692709133396e-06,
"loss": 0.194149649143219,
"step": 2020
},
{
"epoch": 0.6496,
"grad_norm": 0.9662922620773315,
"learning_rate": 1.6513684706109311e-06,
"loss": 0.19424891471862793,
"step": 2030
},
{
"epoch": 0.6528,
"grad_norm": 0.8611494302749634,
"learning_rate": 1.6251501534997529e-06,
"loss": 0.19236912727355956,
"step": 2040
},
{
"epoch": 0.656,
"grad_norm": 0.8275763988494873,
"learning_rate": 1.5990410302226405e-06,
"loss": 0.18528724908828736,
"step": 2050
},
{
"epoch": 0.6592,
"grad_norm": 0.9188935160636902,
"learning_rate": 1.5730443595734162e-06,
"loss": 0.19188053607940675,
"step": 2060
},
{
"epoch": 0.6624,
"grad_norm": 1.038678526878357,
"learning_rate": 1.5471633863101982e-06,
"loss": 0.1937857151031494,
"step": 2070
},
{
"epoch": 0.6656,
"grad_norm": 0.9681159853935242,
"learning_rate": 1.521401340750407e-06,
"loss": 0.1949242353439331,
"step": 2080
},
{
"epoch": 0.6688,
"grad_norm": 0.8842741250991821,
"learning_rate": 1.495761438367577e-06,
"loss": 0.1935848593711853,
"step": 2090
},
{
"epoch": 0.672,
"grad_norm": 1.0433149337768555,
"learning_rate": 1.4702468793900187e-06,
"loss": 0.19175238609313966,
"step": 2100
},
{
"epoch": 0.6752,
"grad_norm": 0.9612663388252258,
"learning_rate": 1.444860848401384e-06,
"loss": 0.19264159202575684,
"step": 2110
},
{
"epoch": 0.6784,
"grad_norm": 0.8770299553871155,
"learning_rate": 1.4196065139431866e-06,
"loss": 0.19034202098846437,
"step": 2120
},
{
"epoch": 0.6816,
"grad_norm": 0.9762235283851624,
"learning_rate": 1.3944870281193178e-06,
"loss": 0.18992135524749756,
"step": 2130
},
{
"epoch": 0.6848,
"grad_norm": 0.8780629634857178,
"learning_rate": 1.3695055262026208e-06,
"loss": 0.1962789535522461,
"step": 2140
},
{
"epoch": 0.688,
"grad_norm": 0.8781450986862183,
"learning_rate": 1.3446651262435679e-06,
"loss": 0.18934234380722045,
"step": 2150
},
{
"epoch": 0.6912,
"grad_norm": 0.9071021676063538,
"learning_rate": 1.3199689286810746e-06,
"loss": 0.19153733253479005,
"step": 2160
},
{
"epoch": 0.6944,
"grad_norm": 0.854719340801239,
"learning_rate": 1.2954200159555294e-06,
"loss": 0.1858367443084717,
"step": 2170
},
{
"epoch": 0.6976,
"grad_norm": 0.9517054557800293,
"learning_rate": 1.2710214521240527e-06,
"loss": 0.18999232053756715,
"step": 2180
},
{
"epoch": 0.7008,
"grad_norm": 0.9305564761161804,
"learning_rate": 1.246776282478063e-06,
"loss": 0.19198548793792725,
"step": 2190
},
{
"epoch": 0.704,
"grad_norm": 0.867964506149292,
"learning_rate": 1.222687533163181e-06,
"loss": 0.19081370830535888,
"step": 2200
},
{
"epoch": 0.7072,
"grad_norm": 1.079858660697937,
"learning_rate": 1.1987582108015228e-06,
"loss": 0.19260694980621337,
"step": 2210
},
{
"epoch": 0.7104,
"grad_norm": 1.0393986701965332,
"learning_rate": 1.1749913021164255e-06,
"loss": 0.192878794670105,
"step": 2220
},
{
"epoch": 0.7136,
"grad_norm": 0.9691563248634338,
"learning_rate": 1.1513897735596702e-06,
"loss": 0.1872938871383667,
"step": 2230
},
{
"epoch": 0.7168,
"grad_norm": 0.9591838121414185,
"learning_rate": 1.127956570941218e-06,
"loss": 0.18445268869400025,
"step": 2240
},
{
"epoch": 0.72,
"grad_norm": 1.0650240182876587,
"learning_rate": 1.104694619061533e-06,
"loss": 0.193602454662323,
"step": 2250
},
{
"epoch": 0.7232,
"grad_norm": 0.9803401827812195,
"learning_rate": 1.0816068213465295e-06,
"loss": 0.19585211277008058,
"step": 2260
},
{
"epoch": 0.7264,
"grad_norm": 0.8571217656135559,
"learning_rate": 1.0586960594851762e-06,
"loss": 0.19123618602752684,
"step": 2270
},
{
"epoch": 0.7296,
"grad_norm": 0.9580956697463989,
"learning_rate": 1.0359651930698217e-06,
"loss": 0.1895804524421692,
"step": 2280
},
{
"epoch": 0.7328,
"grad_norm": 0.9303188323974609,
"learning_rate": 1.0134170592392837e-06,
"loss": 0.1917721748352051,
"step": 2290
},
{
"epoch": 0.736,
"grad_norm": 0.9539825320243835,
"learning_rate": 9.910544723247204e-07,
"loss": 0.1903479814529419,
"step": 2300
},
{
"epoch": 0.7392,
"grad_norm": 0.905822217464447,
"learning_rate": 9.688802234983706e-07,
"loss": 0.1895788311958313,
"step": 2310
},
{
"epoch": 0.7424,
"grad_norm": 1.0421855449676514,
"learning_rate": 9.468970804251742e-07,
"loss": 0.19260928630828858,
"step": 2320
},
{
"epoch": 0.7456,
"grad_norm": 0.938606321811676,
"learning_rate": 9.251077869173244e-07,
"loss": 0.18598917722702027,
"step": 2330
},
{
"epoch": 0.7488,
"grad_norm": 0.9746700525283813,
"learning_rate": 9.035150625918054e-07,
"loss": 0.18742547035217286,
"step": 2340
},
{
"epoch": 0.752,
"grad_norm": 0.9145529270172119,
"learning_rate": 8.821216025309395e-07,
"loss": 0.19074430465698242,
"step": 2350
},
{
"epoch": 0.7552,
"grad_norm": 0.8072162866592407,
"learning_rate": 8.609300769460055e-07,
"loss": 0.18499974012374878,
"step": 2360
},
{
"epoch": 0.7584,
"grad_norm": 0.9440052509307861,
"learning_rate": 8.399431308439592e-07,
"loss": 0.19107134342193605,
"step": 2370
},
{
"epoch": 0.7616,
"grad_norm": 0.9550809264183044,
"learning_rate": 8.191633836972962e-07,
"loss": 0.1914163589477539,
"step": 2380
},
{
"epoch": 0.7648,
"grad_norm": 0.8588680624961853,
"learning_rate": 7.985934291171024e-07,
"loss": 0.18734397888183593,
"step": 2390
},
{
"epoch": 0.768,
"grad_norm": 0.9598610401153564,
"learning_rate": 7.7823583452934e-07,
"loss": 0.1909665584564209,
"step": 2400
},
{
"epoch": 0.7712,
"grad_norm": 0.9106743931770325,
"learning_rate": 7.58093140854389e-07,
"loss": 0.19062764644622804,
"step": 2410
},
{
"epoch": 0.7744,
"grad_norm": 0.9662719964981079,
"learning_rate": 7.381678621899077e-07,
"loss": 0.19155207872390748,
"step": 2420
},
{
"epoch": 0.7776,
"grad_norm": 1.0235718488693237,
"learning_rate": 7.184624854970379e-07,
"loss": 0.18778762817382813,
"step": 2430
},
{
"epoch": 0.7808,
"grad_norm": 0.8647546768188477,
"learning_rate": 6.989794702899932e-07,
"loss": 0.1867724895477295,
"step": 2440
},
{
"epoch": 0.784,
"grad_norm": 0.9516528844833374,
"learning_rate": 6.797212483290777e-07,
"loss": 0.18868536949157716,
"step": 2450
},
{
"epoch": 0.7872,
"grad_norm": 0.9623562097549438,
"learning_rate": 6.60690223317171e-07,
"loss": 0.18562084436416626,
"step": 2460
},
{
"epoch": 0.7904,
"grad_norm": 0.8501712679862976,
"learning_rate": 6.418887705997046e-07,
"loss": 0.18841230869293213,
"step": 2470
},
{
"epoch": 0.7936,
"grad_norm": 1.1048811674118042,
"learning_rate": 6.23319236868189e-07,
"loss": 0.1919555187225342,
"step": 2480
},
{
"epoch": 0.7968,
"grad_norm": 0.8838031888008118,
"learning_rate": 6.049839398673141e-07,
"loss": 0.19270492792129518,
"step": 2490
},
{
"epoch": 0.8,
"grad_norm": 0.9717556238174438,
"learning_rate": 5.868851681056567e-07,
"loss": 0.19487048387527467,
"step": 2500
},
{
"epoch": 0.8032,
"grad_norm": 0.9107718467712402,
"learning_rate": 5.690251805700467e-07,
"loss": 0.1918075442314148,
"step": 2510
},
{
"epoch": 0.8064,
"grad_norm": 0.9718947410583496,
"learning_rate": 5.514062064436096e-07,
"loss": 0.19251750707626342,
"step": 2520
},
{
"epoch": 0.8096,
"grad_norm": 0.9999333620071411,
"learning_rate": 5.34030444827533e-07,
"loss": 0.1888274908065796,
"step": 2530
},
{
"epoch": 0.8128,
"grad_norm": 0.9839712381362915,
"learning_rate": 5.169000644665895e-07,
"loss": 0.1859964370727539,
"step": 2540
},
{
"epoch": 0.816,
"grad_norm": 0.8771271705627441,
"learning_rate": 5.000172034784442e-07,
"loss": 0.1878008484840393,
"step": 2550
},
{
"epoch": 0.8192,
"grad_norm": 1.0618975162506104,
"learning_rate": 4.833839690867853e-07,
"loss": 0.19025259017944335,
"step": 2560
},
{
"epoch": 0.8224,
"grad_norm": 0.9912660121917725,
"learning_rate": 4.6700243735831705e-07,
"loss": 0.18360999822616578,
"step": 2570
},
{
"epoch": 0.8256,
"grad_norm": 0.9649500250816345,
"learning_rate": 4.508746529436311e-07,
"loss": 0.18223977088928223,
"step": 2580
},
{
"epoch": 0.8288,
"grad_norm": 0.9794525504112244,
"learning_rate": 4.350026288220083e-07,
"loss": 0.19027379751205445,
"step": 2590
},
{
"epoch": 0.832,
"grad_norm": 0.9974610209465027,
"learning_rate": 4.1938834605017133e-07,
"loss": 0.1869723081588745,
"step": 2600
},
{
"epoch": 0.8352,
"grad_norm": 0.9569944143295288,
"learning_rate": 4.0403375351501515e-07,
"loss": 0.18647921085357666,
"step": 2610
},
{
"epoch": 0.8384,
"grad_norm": 0.8668234944343567,
"learning_rate": 3.88940767690362e-07,
"loss": 0.18584033250808715,
"step": 2620
},
{
"epoch": 0.8416,
"grad_norm": 1.0080299377441406,
"learning_rate": 3.7411127239775774e-07,
"loss": 0.1850695490837097,
"step": 2630
},
{
"epoch": 0.8448,
"grad_norm": 0.9814310669898987,
"learning_rate": 3.595471185713431e-07,
"loss": 0.18304891586303712,
"step": 2640
},
{
"epoch": 0.848,
"grad_norm": 0.9878867268562317,
"learning_rate": 3.4525012402682826e-07,
"loss": 0.18500676155090331,
"step": 2650
},
{
"epoch": 0.8512,
"grad_norm": 1.0064340829849243,
"learning_rate": 3.3122207323460804e-07,
"loss": 0.18739557266235352,
"step": 2660
},
{
"epoch": 0.8544,
"grad_norm": 0.9369391798973083,
"learning_rate": 3.1746471709702963e-07,
"loss": 0.18277575969696044,
"step": 2670
},
{
"epoch": 0.8576,
"grad_norm": 0.9446983337402344,
"learning_rate": 3.039797727298585e-07,
"loss": 0.1903679609298706,
"step": 2680
},
{
"epoch": 0.8608,
"grad_norm": 1.0289263725280762,
"learning_rate": 2.9076892324795546e-07,
"loss": 0.18804672956466675,
"step": 2690
},
{
"epoch": 0.864,
"grad_norm": 1.047361135482788,
"learning_rate": 2.778338175551995e-07,
"loss": 0.18376014232635499,
"step": 2700
},
{
"epoch": 0.8672,
"grad_norm": 1.0056003332138062,
"learning_rate": 2.6517607013868326e-07,
"loss": 0.19209181070327758,
"step": 2710
},
{
"epoch": 0.8704,
"grad_norm": 0.9757674336433411,
"learning_rate": 2.527972608672002e-07,
"loss": 0.18722193241119384,
"step": 2720
},
{
"epoch": 0.8736,
"grad_norm": 1.0518726110458374,
"learning_rate": 2.40698934794053e-07,
"loss": 0.187286376953125,
"step": 2730
},
{
"epoch": 0.8768,
"grad_norm": 0.93593430519104,
"learning_rate": 2.2888260196421237e-07,
"loss": 0.1866356611251831,
"step": 2740
},
{
"epoch": 0.88,
"grad_norm": 0.9336698651313782,
"learning_rate": 2.1734973722583735e-07,
"loss": 0.1894855499267578,
"step": 2750
},
{
"epoch": 0.8832,
"grad_norm": 1.0136207342147827,
"learning_rate": 2.0610178004619564e-07,
"loss": 0.1810680389404297,
"step": 2760
},
{
"epoch": 0.8864,
"grad_norm": 0.8792027831077576,
"learning_rate": 1.9514013433199834e-07,
"loss": 0.1932243824005127,
"step": 2770
},
{
"epoch": 0.8896,
"grad_norm": 0.8541420698165894,
"learning_rate": 1.8446616825416958e-07,
"loss": 0.1917858123779297,
"step": 2780
},
{
"epoch": 0.8928,
"grad_norm": 1.0666835308074951,
"learning_rate": 1.7408121407708007e-07,
"loss": 0.1921330451965332,
"step": 2790
},
{
"epoch": 0.896,
"grad_norm": 1.0335843563079834,
"learning_rate": 1.6398656799226253e-07,
"loss": 0.1800443172454834,
"step": 2800
},
{
"epoch": 0.8992,
"grad_norm": 1.0246766805648804,
"learning_rate": 1.5418348995662773e-07,
"loss": 0.18684718608856202,
"step": 2810
},
{
"epoch": 0.9024,
"grad_norm": 1.0253738164901733,
"learning_rate": 1.4467320353520275e-07,
"loss": 0.1846174955368042,
"step": 2820
},
{
"epoch": 0.9056,
"grad_norm": 0.8940634727478027,
"learning_rate": 1.3545689574841341e-07,
"loss": 0.1857313632965088,
"step": 2830
},
{
"epoch": 0.9088,
"grad_norm": 0.9817743897438049,
"learning_rate": 1.26535716923927e-07,
"loss": 0.19169492721557618,
"step": 2840
},
{
"epoch": 0.912,
"grad_norm": 0.9626559019088745,
"learning_rate": 1.1791078055307493e-07,
"loss": 0.18825350999832152,
"step": 2850
},
{
"epoch": 0.9152,
"grad_norm": 1.0254485607147217,
"learning_rate": 1.0958316315187289e-07,
"loss": 0.1878815174102783,
"step": 2860
},
{
"epoch": 0.9184,
"grad_norm": 0.9539675712585449,
"learning_rate": 1.0155390412665528e-07,
"loss": 0.18591114282608032,
"step": 2870
},
{
"epoch": 0.9216,
"grad_norm": 0.9155623316764832,
"learning_rate": 9.38240056443443e-08,
"loss": 0.18296668529510499,
"step": 2880
},
{
"epoch": 0.9248,
"grad_norm": 0.9511232376098633,
"learning_rate": 8.639443250736402e-08,
"loss": 0.18568153381347657,
"step": 2890
},
{
"epoch": 0.928,
"grad_norm": 0.921597957611084,
"learning_rate": 7.926611203321777e-08,
"loss": 0.18664343357086183,
"step": 2900
},
{
"epoch": 0.9312,
"grad_norm": 0.9718701243400574,
"learning_rate": 7.243993393874882e-08,
"loss": 0.1875317335128784,
"step": 2910
},
{
"epoch": 0.9344,
"grad_norm": 0.9737975597381592,
"learning_rate": 6.591675022908805e-08,
"loss": 0.18509577512741088,
"step": 2920
},
{
"epoch": 0.9376,
"grad_norm": 0.9686116576194763,
"learning_rate": 5.969737509131241e-08,
"loss": 0.18223493099212645,
"step": 2930
},
{
"epoch": 0.9408,
"grad_norm": 0.9519445896148682,
"learning_rate": 5.3782584792823334e-08,
"loss": 0.1867220640182495,
"step": 2940
},
{
"epoch": 0.944,
"grad_norm": 1.0010050535202026,
"learning_rate": 4.817311758445686e-08,
"loss": 0.18873190879821777,
"step": 2950
},
{
"epoch": 0.9472,
"grad_norm": 1.0026582479476929,
"learning_rate": 4.286967360833866e-08,
"loss": 0.1890841007232666,
"step": 2960
},
{
"epoch": 0.9504,
"grad_norm": 0.9423681497573853,
"learning_rate": 3.787291481049754e-08,
"loss": 0.18887195587158204,
"step": 2970
},
{
"epoch": 0.9536,
"grad_norm": 0.984921932220459,
"learning_rate": 3.3183464858244364e-08,
"loss": 0.19451572895050048,
"step": 2980
},
{
"epoch": 0.9568,
"grad_norm": 0.9740422964096069,
"learning_rate": 2.8801909062328992e-08,
"loss": 0.18073530197143556,
"step": 2990
},
{
"epoch": 0.96,
"grad_norm": 0.9824504256248474,
"learning_rate": 2.4728794303886248e-08,
"loss": 0.17942516803741454,
"step": 3000
},
{
"epoch": 0.9632,
"grad_norm": 1.0434110164642334,
"learning_rate": 2.0964628966175794e-08,
"loss": 0.18552230596542357,
"step": 3010
},
{
"epoch": 0.9664,
"grad_norm": 0.9475349187850952,
"learning_rate": 1.750988287113009e-08,
"loss": 0.1841592311859131,
"step": 3020
},
{
"epoch": 0.9696,
"grad_norm": 0.928511917591095,
"learning_rate": 1.4364987220713278e-08,
"loss": 0.19167165756225585,
"step": 3030
},
{
"epoch": 0.9728,
"grad_norm": 0.9599359631538391,
"learning_rate": 1.1530334543099763e-08,
"loss": 0.1884666085243225,
"step": 3040
},
{
"epoch": 0.976,
"grad_norm": 0.9781372547149658,
"learning_rate": 9.006278643683697e-09,
"loss": 0.19168753623962403,
"step": 3050
},
{
"epoch": 0.9792,
"grad_norm": 0.9357272982597351,
"learning_rate": 6.793134560916514e-09,
"loss": 0.1932600259780884,
"step": 3060
},
{
"epoch": 0.9824,
"grad_norm": 0.9409093260765076,
"learning_rate": 4.891178526986451e-09,
"loss": 0.18934156894683837,
"step": 3070
},
{
"epoch": 0.9856,
"grad_norm": 1.0096839666366577,
"learning_rate": 3.3006479333413943e-09,
"loss": 0.19215707778930663,
"step": 3080
},
{
"epoch": 0.9888,
"grad_norm": 0.9278941750526428,
"learning_rate": 2.021741301058422e-09,
"loss": 0.1875869631767273,
"step": 3090
},
{
"epoch": 0.992,
"grad_norm": 0.9983802437782288,
"learning_rate": 1.0546182560652872e-09,
"loss": 0.18937885761260986,
"step": 3100
},
{
"epoch": 0.9952,
"grad_norm": 0.9678918123245239,
"learning_rate": 3.9939950921774607e-10,
"loss": 0.1848172664642334,
"step": 3110
},
{
"epoch": 0.9984,
"grad_norm": 0.97052401304245,
"learning_rate": 5.616684123160854e-11,
"loss": 0.18443670272827148,
"step": 3120
},
{
"epoch": 1.0,
"step": 3125,
"total_flos": 1.4635172064982467e+18,
"train_loss": 0.20465421264648437,
"train_runtime": 15978.2192,
"train_samples_per_second": 12.517,
"train_steps_per_second": 0.196
}
],
"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": 1.4635172064982467e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}