Safetensors
CosFly-Track / checkpoints /Qwen3.5-2B-SFT /trainer_state.json
renruilong's picture
Add scripts and checkpoints (CosFly-Track release) (#85)
284878d
raw
history blame
55.4 kB
{
"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": 17.00181007385254,
"learning_rate": 1.437699680511182e-07,
"loss": 0.40097870826721194,
"step": 10
},
{
"epoch": 0.0064,
"grad_norm": 9.006903648376465,
"learning_rate": 3.0351437699680514e-07,
"loss": 0.3749201536178589,
"step": 20
},
{
"epoch": 0.0096,
"grad_norm": 2.7475762367248535,
"learning_rate": 4.6325878594249205e-07,
"loss": 0.3104386329650879,
"step": 30
},
{
"epoch": 0.0128,
"grad_norm": 1.6237599849700928,
"learning_rate": 6.230031948881789e-07,
"loss": 0.28350210189819336,
"step": 40
},
{
"epoch": 0.016,
"grad_norm": 1.1944012641906738,
"learning_rate": 7.82747603833866e-07,
"loss": 0.26195507049560546,
"step": 50
},
{
"epoch": 0.0192,
"grad_norm": 1.111968994140625,
"learning_rate": 9.424920127795528e-07,
"loss": 0.2548145532608032,
"step": 60
},
{
"epoch": 0.0224,
"grad_norm": 1.0771538019180298,
"learning_rate": 1.1022364217252397e-06,
"loss": 0.26291556358337403,
"step": 70
},
{
"epoch": 0.0256,
"grad_norm": 0.9149865508079529,
"learning_rate": 1.2619808306709266e-06,
"loss": 0.24681921005249025,
"step": 80
},
{
"epoch": 0.0288,
"grad_norm": 1.1506657600402832,
"learning_rate": 1.4217252396166134e-06,
"loss": 0.24312305450439453,
"step": 90
},
{
"epoch": 0.032,
"grad_norm": 1.067981243133545,
"learning_rate": 1.5814696485623005e-06,
"loss": 0.24636492729187012,
"step": 100
},
{
"epoch": 0.0352,
"grad_norm": 1.1217856407165527,
"learning_rate": 1.7412140575079875e-06,
"loss": 0.24657981395721434,
"step": 110
},
{
"epoch": 0.0384,
"grad_norm": 1.265798807144165,
"learning_rate": 1.9009584664536742e-06,
"loss": 0.23692235946655274,
"step": 120
},
{
"epoch": 0.0416,
"grad_norm": 1.001699686050415,
"learning_rate": 2.060702875399361e-06,
"loss": 0.2406609535217285,
"step": 130
},
{
"epoch": 0.0448,
"grad_norm": 1.0411975383758545,
"learning_rate": 2.220447284345048e-06,
"loss": 0.23273375034332275,
"step": 140
},
{
"epoch": 0.048,
"grad_norm": 1.1774442195892334,
"learning_rate": 2.380191693290735e-06,
"loss": 0.23040361404418946,
"step": 150
},
{
"epoch": 0.0512,
"grad_norm": 0.9190797805786133,
"learning_rate": 2.539936102236422e-06,
"loss": 0.2255420684814453,
"step": 160
},
{
"epoch": 0.0544,
"grad_norm": 0.917509913444519,
"learning_rate": 2.699680511182109e-06,
"loss": 0.23383898735046388,
"step": 170
},
{
"epoch": 0.0576,
"grad_norm": 1.1193501949310303,
"learning_rate": 2.8594249201277955e-06,
"loss": 0.22656521797180176,
"step": 180
},
{
"epoch": 0.0608,
"grad_norm": 1.022067904472351,
"learning_rate": 3.0191693290734825e-06,
"loss": 0.23097100257873535,
"step": 190
},
{
"epoch": 0.064,
"grad_norm": 0.9210631847381592,
"learning_rate": 3.17891373801917e-06,
"loss": 0.2329324722290039,
"step": 200
},
{
"epoch": 0.0672,
"grad_norm": 0.9588786959648132,
"learning_rate": 3.3386581469648564e-06,
"loss": 0.2294376850128174,
"step": 210
},
{
"epoch": 0.0704,
"grad_norm": 0.8580502867698669,
"learning_rate": 3.4984025559105434e-06,
"loss": 0.22389583587646483,
"step": 220
},
{
"epoch": 0.0736,
"grad_norm": 0.946774423122406,
"learning_rate": 3.6581469648562303e-06,
"loss": 0.22544665336608888,
"step": 230
},
{
"epoch": 0.0768,
"grad_norm": 0.9038301110267639,
"learning_rate": 3.817891373801918e-06,
"loss": 0.2216867446899414,
"step": 240
},
{
"epoch": 0.08,
"grad_norm": 0.7587664127349854,
"learning_rate": 3.977635782747604e-06,
"loss": 0.2192216396331787,
"step": 250
},
{
"epoch": 0.0832,
"grad_norm": 0.8255429267883301,
"learning_rate": 4.137380191693291e-06,
"loss": 0.22012462615966796,
"step": 260
},
{
"epoch": 0.0864,
"grad_norm": 0.7769321799278259,
"learning_rate": 4.297124600638978e-06,
"loss": 0.22262439727783204,
"step": 270
},
{
"epoch": 0.0896,
"grad_norm": 0.8024886846542358,
"learning_rate": 4.456869009584665e-06,
"loss": 0.21874587535858153,
"step": 280
},
{
"epoch": 0.0928,
"grad_norm": 0.8127010464668274,
"learning_rate": 4.616613418530352e-06,
"loss": 0.22413947582244872,
"step": 290
},
{
"epoch": 0.096,
"grad_norm": 0.8755475878715515,
"learning_rate": 4.776357827476039e-06,
"loss": 0.2250971794128418,
"step": 300
},
{
"epoch": 0.0992,
"grad_norm": 0.9584972858428955,
"learning_rate": 4.936102236421725e-06,
"loss": 0.21371636390686036,
"step": 310
},
{
"epoch": 0.1024,
"grad_norm": 0.8256917595863342,
"learning_rate": 4.999943833158769e-06,
"loss": 0.21431188583374022,
"step": 320
},
{
"epoch": 0.1056,
"grad_norm": 0.7529917359352112,
"learning_rate": 4.999600600490783e-06,
"loss": 0.22084765434265136,
"step": 330
},
{
"epoch": 0.1088,
"grad_norm": 0.7913526296615601,
"learning_rate": 4.9989453817439345e-06,
"loss": 0.21453213691711426,
"step": 340
},
{
"epoch": 0.112,
"grad_norm": 0.7930585145950317,
"learning_rate": 4.997978258698942e-06,
"loss": 0.21424226760864257,
"step": 350
},
{
"epoch": 0.1152,
"grad_norm": 0.7409906983375549,
"learning_rate": 4.996699352066659e-06,
"loss": 0.21544296741485597,
"step": 360
},
{
"epoch": 0.1184,
"grad_norm": 0.776908814907074,
"learning_rate": 4.995108821473014e-06,
"loss": 0.21555554866790771,
"step": 370
},
{
"epoch": 0.1216,
"grad_norm": 0.7414209842681885,
"learning_rate": 4.993206865439084e-06,
"loss": 0.21163430213928222,
"step": 380
},
{
"epoch": 0.1248,
"grad_norm": 0.7566289901733398,
"learning_rate": 4.990993721356317e-06,
"loss": 0.21131091117858886,
"step": 390
},
{
"epoch": 0.128,
"grad_norm": 0.7883878946304321,
"learning_rate": 4.988469665456901e-06,
"loss": 0.2116244316101074,
"step": 400
},
{
"epoch": 0.1312,
"grad_norm": 0.7117395401000977,
"learning_rate": 4.985635012779288e-06,
"loss": 0.21834149360656738,
"step": 410
},
{
"epoch": 0.1344,
"grad_norm": 0.6810120940208435,
"learning_rate": 4.98249011712887e-06,
"loss": 0.21092677116394043,
"step": 420
},
{
"epoch": 0.1376,
"grad_norm": 0.7290633320808411,
"learning_rate": 4.979035371033824e-06,
"loss": 0.2102036714553833,
"step": 430
},
{
"epoch": 0.1408,
"grad_norm": 0.7017252445220947,
"learning_rate": 4.975271205696115e-06,
"loss": 0.20992655754089357,
"step": 440
},
{
"epoch": 0.144,
"grad_norm": 0.7360793948173523,
"learning_rate": 4.971198090937671e-06,
"loss": 0.20512619018554687,
"step": 450
},
{
"epoch": 0.1472,
"grad_norm": 0.7540563344955444,
"learning_rate": 4.966816535141756e-06,
"loss": 0.2015425443649292,
"step": 460
},
{
"epoch": 0.1504,
"grad_norm": 0.7753161787986755,
"learning_rate": 4.9621270851895035e-06,
"loss": 0.21057233810424805,
"step": 470
},
{
"epoch": 0.1536,
"grad_norm": 0.7387011647224426,
"learning_rate": 4.957130326391662e-06,
"loss": 0.21185002326965333,
"step": 480
},
{
"epoch": 0.1568,
"grad_norm": 0.7276087999343872,
"learning_rate": 4.951826882415544e-06,
"loss": 0.20872690677642822,
"step": 490
},
{
"epoch": 0.16,
"grad_norm": 0.8299422264099121,
"learning_rate": 4.946217415207177e-06,
"loss": 0.20370903015136718,
"step": 500
},
{
"epoch": 0.1632,
"grad_norm": 0.7076508402824402,
"learning_rate": 4.940302624908689e-06,
"loss": 0.20804412364959718,
"step": 510
},
{
"epoch": 0.1664,
"grad_norm": 0.706371545791626,
"learning_rate": 4.934083249770912e-06,
"loss": 0.20231704711914061,
"step": 520
},
{
"epoch": 0.1696,
"grad_norm": 0.7540279030799866,
"learning_rate": 4.927560066061251e-06,
"loss": 0.2063138246536255,
"step": 530
},
{
"epoch": 0.1728,
"grad_norm": 0.6493498682975769,
"learning_rate": 4.920733887966783e-06,
"loss": 0.21687688827514648,
"step": 540
},
{
"epoch": 0.176,
"grad_norm": 0.6475550532341003,
"learning_rate": 4.913605567492636e-06,
"loss": 0.2051780939102173,
"step": 550
},
{
"epoch": 0.1792,
"grad_norm": 0.7162631750106812,
"learning_rate": 4.906175994355656e-06,
"loss": 0.20715060234069824,
"step": 560
},
{
"epoch": 0.1824,
"grad_norm": 0.8360611796379089,
"learning_rate": 4.898446095873345e-06,
"loss": 0.20873281955718995,
"step": 570
},
{
"epoch": 0.1856,
"grad_norm": 0.615298867225647,
"learning_rate": 4.890416836848128e-06,
"loss": 0.20635380744934081,
"step": 580
},
{
"epoch": 0.1888,
"grad_norm": 0.7096477746963501,
"learning_rate": 4.882089219446925e-06,
"loss": 0.20148007869720458,
"step": 590
},
{
"epoch": 0.192,
"grad_norm": 0.7298900485038757,
"learning_rate": 4.873464283076074e-06,
"loss": 0.2067399024963379,
"step": 600
},
{
"epoch": 0.1952,
"grad_norm": 0.7385826706886292,
"learning_rate": 4.864543104251587e-06,
"loss": 0.20580265522003174,
"step": 610
},
{
"epoch": 0.1984,
"grad_norm": 0.65291827917099,
"learning_rate": 4.855326796464798e-06,
"loss": 0.2078333854675293,
"step": 620
},
{
"epoch": 0.2016,
"grad_norm": 0.6820300817489624,
"learning_rate": 4.8458165100433725e-06,
"loss": 0.20342307090759276,
"step": 630
},
{
"epoch": 0.2048,
"grad_norm": 0.7165417075157166,
"learning_rate": 4.836013432007738e-06,
"loss": 0.20381028652191163,
"step": 640
},
{
"epoch": 0.208,
"grad_norm": 0.7687081694602966,
"learning_rate": 4.825918785922921e-06,
"loss": 0.2078242301940918,
"step": 650
},
{
"epoch": 0.2112,
"grad_norm": 0.7032524347305298,
"learning_rate": 4.8155338317458315e-06,
"loss": 0.20481655597686768,
"step": 660
},
{
"epoch": 0.2144,
"grad_norm": 0.7231862545013428,
"learning_rate": 4.804859865668002e-06,
"loss": 0.20165665149688722,
"step": 670
},
{
"epoch": 0.2176,
"grad_norm": 0.6389712691307068,
"learning_rate": 4.793898219953804e-06,
"loss": 0.20336999893188476,
"step": 680
},
{
"epoch": 0.2208,
"grad_norm": 0.6152437925338745,
"learning_rate": 4.782650262774164e-06,
"loss": 0.20431649684906006,
"step": 690
},
{
"epoch": 0.224,
"grad_norm": 0.6727204918861389,
"learning_rate": 4.7711173980357886e-06,
"loss": 0.2011037588119507,
"step": 700
},
{
"epoch": 0.2272,
"grad_norm": 0.6739911437034607,
"learning_rate": 4.759301065205947e-06,
"loss": 0.20095202922821045,
"step": 710
},
{
"epoch": 0.2304,
"grad_norm": 0.7094120979309082,
"learning_rate": 4.7472027391328e-06,
"loss": 0.20286693572998046,
"step": 720
},
{
"epoch": 0.2336,
"grad_norm": 0.7743499875068665,
"learning_rate": 4.734823929861317e-06,
"loss": 0.20463786125183106,
"step": 730
},
{
"epoch": 0.2368,
"grad_norm": 0.8282691836357117,
"learning_rate": 4.722166182444801e-06,
"loss": 0.20196740627288817,
"step": 740
},
{
"epoch": 0.24,
"grad_norm": 0.7588798403739929,
"learning_rate": 4.709231076752045e-06,
"loss": 0.20084238052368164,
"step": 750
},
{
"epoch": 0.2432,
"grad_norm": 0.6279271245002747,
"learning_rate": 4.696020227270142e-06,
"loss": 0.20528481006622315,
"step": 760
},
{
"epoch": 0.2464,
"grad_norm": 0.6513218879699707,
"learning_rate": 4.6825352829029705e-06,
"loss": 0.200866436958313,
"step": 770
},
{
"epoch": 0.2496,
"grad_norm": 0.6800708174705505,
"learning_rate": 4.668777926765392e-06,
"loss": 0.2006533145904541,
"step": 780
},
{
"epoch": 0.2528,
"grad_norm": 0.678886890411377,
"learning_rate": 4.6547498759731725e-06,
"loss": 0.1949576735496521,
"step": 790
},
{
"epoch": 0.256,
"grad_norm": 0.7107610702514648,
"learning_rate": 4.6404528814286575e-06,
"loss": 0.19777085781097412,
"step": 800
},
{
"epoch": 0.2592,
"grad_norm": 0.6964312791824341,
"learning_rate": 4.6258887276022425e-06,
"loss": 0.2050178050994873,
"step": 810
},
{
"epoch": 0.2624,
"grad_norm": 0.725965142250061,
"learning_rate": 4.611059232309639e-06,
"loss": 0.20062518119812012,
"step": 820
},
{
"epoch": 0.2656,
"grad_norm": 0.7800929546356201,
"learning_rate": 4.595966246484986e-06,
"loss": 0.20146868228912354,
"step": 830
},
{
"epoch": 0.2688,
"grad_norm": 0.749195396900177,
"learning_rate": 4.580611653949829e-06,
"loss": 0.20182161331176757,
"step": 840
},
{
"epoch": 0.272,
"grad_norm": 0.7181940078735352,
"learning_rate": 4.564997371177992e-06,
"loss": 0.1992729663848877,
"step": 850
},
{
"epoch": 0.2752,
"grad_norm": 0.6476516723632812,
"learning_rate": 4.54912534705637e-06,
"loss": 0.19952178001403809,
"step": 860
},
{
"epoch": 0.2784,
"grad_norm": 0.6579780578613281,
"learning_rate": 4.532997562641683e-06,
"loss": 0.19619479179382324,
"step": 870
},
{
"epoch": 0.2816,
"grad_norm": 0.6974263787269592,
"learning_rate": 4.516616030913214e-06,
"loss": 0.20040647983551024,
"step": 880
},
{
"epoch": 0.2848,
"grad_norm": 0.6694504618644714,
"learning_rate": 4.499982796521556e-06,
"loss": 0.19575827121734618,
"step": 890
},
{
"epoch": 0.288,
"grad_norm": 0.6807686686515808,
"learning_rate": 4.48309993553341e-06,
"loss": 0.19785771369934083,
"step": 900
},
{
"epoch": 0.2912,
"grad_norm": 0.6845031976699829,
"learning_rate": 4.465969555172468e-06,
"loss": 0.19707002639770507,
"step": 910
},
{
"epoch": 0.2944,
"grad_norm": 0.7153148055076599,
"learning_rate": 4.448593793556391e-06,
"loss": 0.2026472806930542,
"step": 920
},
{
"epoch": 0.2976,
"grad_norm": 0.6482206583023071,
"learning_rate": 4.430974819429954e-06,
"loss": 0.19655225276947022,
"step": 930
},
{
"epoch": 0.3008,
"grad_norm": 0.6306204795837402,
"learning_rate": 4.413114831894344e-06,
"loss": 0.19124293327331543,
"step": 940
},
{
"epoch": 0.304,
"grad_norm": 0.6225886940956116,
"learning_rate": 4.3950160601326865e-06,
"loss": 0.19019542932510375,
"step": 950
},
{
"epoch": 0.3072,
"grad_norm": 0.7444518804550171,
"learning_rate": 4.376680763131811e-06,
"loss": 0.1960599660873413,
"step": 960
},
{
"epoch": 0.3104,
"grad_norm": 0.7806506752967834,
"learning_rate": 4.358111229400296e-06,
"loss": 0.19993667602539061,
"step": 970
},
{
"epoch": 0.3136,
"grad_norm": 0.7076775431632996,
"learning_rate": 4.33930977668283e-06,
"loss": 0.20315132141113282,
"step": 980
},
{
"epoch": 0.3168,
"grad_norm": 0.681239664554596,
"learning_rate": 4.320278751670922e-06,
"loss": 0.19584376811981202,
"step": 990
},
{
"epoch": 0.32,
"grad_norm": 0.7305501103401184,
"learning_rate": 4.301020529710009e-06,
"loss": 0.19784679412841796,
"step": 1000
},
{
"epoch": 0.3232,
"grad_norm": 0.7520341873168945,
"learning_rate": 4.281537514502962e-06,
"loss": 0.20119853019714357,
"step": 1010
},
{
"epoch": 0.3264,
"grad_norm": 0.6639211177825928,
"learning_rate": 4.261832137810093e-06,
"loss": 0.197571325302124,
"step": 1020
},
{
"epoch": 0.3296,
"grad_norm": 0.6915456056594849,
"learning_rate": 4.241906859145611e-06,
"loss": 0.19875795841217042,
"step": 1030
},
{
"epoch": 0.3328,
"grad_norm": 0.6685866117477417,
"learning_rate": 4.221764165470661e-06,
"loss": 0.19526104927062987,
"step": 1040
},
{
"epoch": 0.336,
"grad_norm": 0.7343277335166931,
"learning_rate": 4.201406570882898e-06,
"loss": 0.19495513439178466,
"step": 1050
},
{
"epoch": 0.3392,
"grad_norm": 0.6586959958076477,
"learning_rate": 4.180836616302704e-06,
"loss": 0.19428329467773436,
"step": 1060
},
{
"epoch": 0.3424,
"grad_norm": 0.7318632006645203,
"learning_rate": 4.160056869156041e-06,
"loss": 0.19764463901519774,
"step": 1070
},
{
"epoch": 0.3456,
"grad_norm": 0.6407229900360107,
"learning_rate": 4.139069923053995e-06,
"loss": 0.19589632749557495,
"step": 1080
},
{
"epoch": 0.3488,
"grad_norm": 0.6162118911743164,
"learning_rate": 4.117878397469062e-06,
"loss": 0.1992954969406128,
"step": 1090
},
{
"epoch": 0.352,
"grad_norm": 0.6185060739517212,
"learning_rate": 4.096484937408195e-06,
"loss": 0.19082267284393312,
"step": 1100
},
{
"epoch": 0.3552,
"grad_norm": 0.6647833585739136,
"learning_rate": 4.074892213082676e-06,
"loss": 0.1906151533126831,
"step": 1110
},
{
"epoch": 0.3584,
"grad_norm": 0.7094317674636841,
"learning_rate": 4.0531029195748265e-06,
"loss": 0.19853758811950684,
"step": 1120
},
{
"epoch": 0.3616,
"grad_norm": 0.6957128643989563,
"learning_rate": 4.03111977650163e-06,
"loss": 0.1966111183166504,
"step": 1130
},
{
"epoch": 0.3648,
"grad_norm": 0.6727164387702942,
"learning_rate": 4.008945527675281e-06,
"loss": 0.19672225713729857,
"step": 1140
},
{
"epoch": 0.368,
"grad_norm": 0.7790160179138184,
"learning_rate": 3.986582940760717e-06,
"loss": 0.18639541864395143,
"step": 1150
},
{
"epoch": 0.3712,
"grad_norm": 0.7126836776733398,
"learning_rate": 3.9640348069301785e-06,
"loss": 0.19289255142211914,
"step": 1160
},
{
"epoch": 0.3744,
"grad_norm": 0.6615356802940369,
"learning_rate": 3.941303940514826e-06,
"loss": 0.19387151002883912,
"step": 1170
},
{
"epoch": 0.3776,
"grad_norm": 0.7060992121696472,
"learning_rate": 3.918393178653472e-06,
"loss": 0.19526903629302977,
"step": 1180
},
{
"epoch": 0.3808,
"grad_norm": 0.6411144137382507,
"learning_rate": 3.895305380938468e-06,
"loss": 0.19223073720932007,
"step": 1190
},
{
"epoch": 0.384,
"grad_norm": 0.6146537065505981,
"learning_rate": 3.872043429058783e-06,
"loss": 0.1911651611328125,
"step": 1200
},
{
"epoch": 0.3872,
"grad_norm": 0.6731158494949341,
"learning_rate": 3.84861022644033e-06,
"loss": 0.1918567180633545,
"step": 1210
},
{
"epoch": 0.3904,
"grad_norm": 0.7398373484611511,
"learning_rate": 3.825008697883574e-06,
"loss": 0.20071697235107422,
"step": 1220
},
{
"epoch": 0.3936,
"grad_norm": 0.7013578414916992,
"learning_rate": 3.8012417891984776e-06,
"loss": 0.19355409145355223,
"step": 1230
},
{
"epoch": 0.3968,
"grad_norm": 0.698662519454956,
"learning_rate": 3.777312466836819e-06,
"loss": 0.19278268814086913,
"step": 1240
},
{
"epoch": 0.4,
"grad_norm": 0.6574701070785522,
"learning_rate": 3.7532237175219378e-06,
"loss": 0.19122891426086425,
"step": 1250
},
{
"epoch": 0.4032,
"grad_norm": 0.7399858832359314,
"learning_rate": 3.728978547875948e-06,
"loss": 0.1955470085144043,
"step": 1260
},
{
"epoch": 0.4064,
"grad_norm": 0.6889676451683044,
"learning_rate": 3.7045799840444712e-06,
"loss": 0.18996728658676149,
"step": 1270
},
{
"epoch": 0.4096,
"grad_norm": 0.73622727394104,
"learning_rate": 3.6800310713189258e-06,
"loss": 0.19117178916931152,
"step": 1280
},
{
"epoch": 0.4128,
"grad_norm": 0.8547033071517944,
"learning_rate": 3.6553348737564328e-06,
"loss": 0.19134058952331542,
"step": 1290
},
{
"epoch": 0.416,
"grad_norm": 0.6093372702598572,
"learning_rate": 3.6304944737973794e-06,
"loss": 0.197102952003479,
"step": 1300
},
{
"epoch": 0.4192,
"grad_norm": 0.7401471138000488,
"learning_rate": 3.6055129718806836e-06,
"loss": 0.19205976724624635,
"step": 1310
},
{
"epoch": 0.4224,
"grad_norm": 0.7101146578788757,
"learning_rate": 3.5803934860568134e-06,
"loss": 0.18738794326782227,
"step": 1320
},
{
"epoch": 0.4256,
"grad_norm": 0.7336035370826721,
"learning_rate": 3.5551391515986163e-06,
"loss": 0.1925289273262024,
"step": 1330
},
{
"epoch": 0.4288,
"grad_norm": 0.6703352928161621,
"learning_rate": 3.529753120609982e-06,
"loss": 0.1879317045211792,
"step": 1340
},
{
"epoch": 0.432,
"grad_norm": 0.7639795541763306,
"learning_rate": 3.5042385616324243e-06,
"loss": 0.19051835536956788,
"step": 1350
},
{
"epoch": 0.4352,
"grad_norm": 0.6993486285209656,
"learning_rate": 3.4785986592495934e-06,
"loss": 0.18969300985336304,
"step": 1360
},
{
"epoch": 0.4384,
"grad_norm": 0.6297987103462219,
"learning_rate": 3.452836613689803e-06,
"loss": 0.1876993417739868,
"step": 1370
},
{
"epoch": 0.4416,
"grad_norm": 0.6074863076210022,
"learning_rate": 3.426955640426584e-06,
"loss": 0.1906881809234619,
"step": 1380
},
{
"epoch": 0.4448,
"grad_norm": 0.6951903700828552,
"learning_rate": 3.4009589697773605e-06,
"loss": 0.18978222608566284,
"step": 1390
},
{
"epoch": 0.448,
"grad_norm": 0.7054743766784668,
"learning_rate": 3.3748498465002475e-06,
"loss": 0.18714208602905275,
"step": 1400
},
{
"epoch": 0.4512,
"grad_norm": 0.6384505033493042,
"learning_rate": 3.3486315293890693e-06,
"loss": 0.19547748565673828,
"step": 1410
},
{
"epoch": 0.4544,
"grad_norm": 0.6575579643249512,
"learning_rate": 3.3223072908666053e-06,
"loss": 0.18565211296081544,
"step": 1420
},
{
"epoch": 0.4576,
"grad_norm": 0.6746217608451843,
"learning_rate": 3.295880416576153e-06,
"loss": 0.1961667060852051,
"step": 1430
},
{
"epoch": 0.4608,
"grad_norm": 0.6876834630966187,
"learning_rate": 3.269354204971427e-06,
"loss": 0.18892121315002441,
"step": 1440
},
{
"epoch": 0.464,
"grad_norm": 0.753948986530304,
"learning_rate": 3.242731966904865e-06,
"loss": 0.186625075340271,
"step": 1450
},
{
"epoch": 0.4672,
"grad_norm": 0.6489424109458923,
"learning_rate": 3.2160170252143913e-06,
"loss": 0.18691904544830323,
"step": 1460
},
{
"epoch": 0.4704,
"grad_norm": 0.619157075881958,
"learning_rate": 3.1892127143086716e-06,
"loss": 0.1932499647140503,
"step": 1470
},
{
"epoch": 0.4736,
"grad_norm": 0.693301796913147,
"learning_rate": 3.1623223797509347e-06,
"loss": 0.18296878337860106,
"step": 1480
},
{
"epoch": 0.4768,
"grad_norm": 0.7320811152458191,
"learning_rate": 3.135349377841396e-06,
"loss": 0.1871619701385498,
"step": 1490
},
{
"epoch": 0.48,
"grad_norm": 0.7020161747932434,
"learning_rate": 3.1082970751983497e-06,
"loss": 0.19298884868621827,
"step": 1500
},
{
"epoch": 0.4832,
"grad_norm": 0.7094489932060242,
"learning_rate": 3.0811688483379546e-06,
"loss": 0.19058284759521485,
"step": 1510
},
{
"epoch": 0.4864,
"grad_norm": 0.6349408030509949,
"learning_rate": 3.0539680832528074e-06,
"loss": 0.19067602157592772,
"step": 1520
},
{
"epoch": 0.4896,
"grad_norm": 0.6655322909355164,
"learning_rate": 3.026698174989316e-06,
"loss": 0.18840067386627196,
"step": 1530
},
{
"epoch": 0.4928,
"grad_norm": 0.6164231896400452,
"learning_rate": 2.999362527223952e-06,
"loss": 0.1893659234046936,
"step": 1540
},
{
"epoch": 0.496,
"grad_norm": 0.6291287541389465,
"learning_rate": 2.9719645518384194e-06,
"loss": 0.19131317138671874,
"step": 1550
},
{
"epoch": 0.4992,
"grad_norm": 0.7157194018363953,
"learning_rate": 2.944507668493807e-06,
"loss": 0.19394657611846924,
"step": 1560
},
{
"epoch": 0.5024,
"grad_norm": 0.701957106590271,
"learning_rate": 2.9169953042037623e-06,
"loss": 0.18929052352905273,
"step": 1570
},
{
"epoch": 0.5056,
"grad_norm": 0.7337380051612854,
"learning_rate": 2.889430892906754e-06,
"loss": 0.1855141282081604,
"step": 1580
},
{
"epoch": 0.5088,
"grad_norm": 0.7762733101844788,
"learning_rate": 2.861817875037462e-06,
"loss": 0.18262506723403932,
"step": 1590
},
{
"epoch": 0.512,
"grad_norm": 0.721386730670929,
"learning_rate": 2.8341596970973683e-06,
"loss": 0.1928829550743103,
"step": 1600
},
{
"epoch": 0.5152,
"grad_norm": 0.6835659146308899,
"learning_rate": 2.80645981122458e-06,
"loss": 0.18776969909667968,
"step": 1610
},
{
"epoch": 0.5184,
"grad_norm": 0.6848421692848206,
"learning_rate": 2.7787216747629508e-06,
"loss": 0.19408185482025148,
"step": 1620
},
{
"epoch": 0.5216,
"grad_norm": 0.6096782684326172,
"learning_rate": 2.7509487498305615e-06,
"loss": 0.18251194953918456,
"step": 1630
},
{
"epoch": 0.5248,
"grad_norm": 0.6688647270202637,
"learning_rate": 2.7231445028875924e-06,
"loss": 0.18787600994110107,
"step": 1640
},
{
"epoch": 0.528,
"grad_norm": 0.6558749675750732,
"learning_rate": 2.6953124043036604e-06,
"loss": 0.18639014959335326,
"step": 1650
},
{
"epoch": 0.5312,
"grad_norm": 0.6501591801643372,
"learning_rate": 2.667455927924667e-06,
"loss": 0.18570780754089355,
"step": 1660
},
{
"epoch": 0.5344,
"grad_norm": 0.6679975986480713,
"learning_rate": 2.6395785506392164e-06,
"loss": 0.18148298263549806,
"step": 1670
},
{
"epoch": 0.5376,
"grad_norm": 0.6896058320999146,
"learning_rate": 2.6116837519446407e-06,
"loss": 0.18533236980438234,
"step": 1680
},
{
"epoch": 0.5408,
"grad_norm": 0.6504887342453003,
"learning_rate": 2.5837750135127192e-06,
"loss": 0.18286749124526977,
"step": 1690
},
{
"epoch": 0.544,
"grad_norm": 0.6970167756080627,
"learning_rate": 2.555855818755108e-06,
"loss": 0.1885154962539673,
"step": 1700
},
{
"epoch": 0.5472,
"grad_norm": 0.7530934810638428,
"learning_rate": 2.5279296523885636e-06,
"loss": 0.18549870252609252,
"step": 1710
},
{
"epoch": 0.5504,
"grad_norm": 0.7539217472076416,
"learning_rate": 2.5e-06,
"loss": 0.18675594329833983,
"step": 1720
},
{
"epoch": 0.5536,
"grad_norm": 0.7183989882469177,
"learning_rate": 2.472070347611437e-06,
"loss": 0.181170654296875,
"step": 1730
},
{
"epoch": 0.5568,
"grad_norm": 0.6666719317436218,
"learning_rate": 2.444144181244893e-06,
"loss": 0.18674240112304688,
"step": 1740
},
{
"epoch": 0.56,
"grad_norm": 0.7493124604225159,
"learning_rate": 2.416224986487282e-06,
"loss": 0.18279486894607544,
"step": 1750
},
{
"epoch": 0.5632,
"grad_norm": 0.683366596698761,
"learning_rate": 2.3883162480553605e-06,
"loss": 0.18034757375717164,
"step": 1760
},
{
"epoch": 0.5664,
"grad_norm": 0.7152907252311707,
"learning_rate": 2.3604214493607844e-06,
"loss": 0.18426483869552612,
"step": 1770
},
{
"epoch": 0.5696,
"grad_norm": 0.7228681445121765,
"learning_rate": 2.332544072075333e-06,
"loss": 0.18772796392440796,
"step": 1780
},
{
"epoch": 0.5728,
"grad_norm": 0.6640520095825195,
"learning_rate": 2.30468759569634e-06,
"loss": 0.18662737607955932,
"step": 1790
},
{
"epoch": 0.576,
"grad_norm": 0.7385361194610596,
"learning_rate": 2.276855497112408e-06,
"loss": 0.18393266201019287,
"step": 1800
},
{
"epoch": 0.5792,
"grad_norm": 0.6731036305427551,
"learning_rate": 2.2490512501694394e-06,
"loss": 0.17882565259933472,
"step": 1810
},
{
"epoch": 0.5824,
"grad_norm": 0.6727462410926819,
"learning_rate": 2.2212783252370496e-06,
"loss": 0.1844714403152466,
"step": 1820
},
{
"epoch": 0.5856,
"grad_norm": 0.6584063768386841,
"learning_rate": 2.1935401887754213e-06,
"loss": 0.1889561891555786,
"step": 1830
},
{
"epoch": 0.5888,
"grad_norm": 0.7410686016082764,
"learning_rate": 2.165840302902632e-06,
"loss": 0.18200514316558838,
"step": 1840
},
{
"epoch": 0.592,
"grad_norm": 0.6837431788444519,
"learning_rate": 2.1381821249625383e-06,
"loss": 0.18779854774475097,
"step": 1850
},
{
"epoch": 0.5952,
"grad_norm": 0.7690724730491638,
"learning_rate": 2.1105691070932465e-06,
"loss": 0.18021690845489502,
"step": 1860
},
{
"epoch": 0.5984,
"grad_norm": 0.7139452695846558,
"learning_rate": 2.083004695796238e-06,
"loss": 0.1776273250579834,
"step": 1870
},
{
"epoch": 0.6016,
"grad_norm": 0.7305198907852173,
"learning_rate": 2.055492331506194e-06,
"loss": 0.18511974811553955,
"step": 1880
},
{
"epoch": 0.6048,
"grad_norm": 0.7747855186462402,
"learning_rate": 2.0280354481615814e-06,
"loss": 0.18330963850021362,
"step": 1890
},
{
"epoch": 0.608,
"grad_norm": 0.7213889360427856,
"learning_rate": 2.000637472776049e-06,
"loss": 0.18462073802947998,
"step": 1900
},
{
"epoch": 0.6112,
"grad_norm": 0.6749533414840698,
"learning_rate": 1.973301825010685e-06,
"loss": 0.17921651601791383,
"step": 1910
},
{
"epoch": 0.6144,
"grad_norm": 0.7167413830757141,
"learning_rate": 1.9460319167471934e-06,
"loss": 0.18418827056884765,
"step": 1920
},
{
"epoch": 0.6176,
"grad_norm": 0.6720898747444153,
"learning_rate": 1.9188311516620466e-06,
"loss": 0.17980489730834961,
"step": 1930
},
{
"epoch": 0.6208,
"grad_norm": 0.721839189529419,
"learning_rate": 1.891702924801651e-06,
"loss": 0.18582551479339598,
"step": 1940
},
{
"epoch": 0.624,
"grad_norm": 0.6840994954109192,
"learning_rate": 1.864650622158604e-06,
"loss": 0.19025051593780518,
"step": 1950
},
{
"epoch": 0.6272,
"grad_norm": 0.649747908115387,
"learning_rate": 1.8376776202490666e-06,
"loss": 0.18501710891723633,
"step": 1960
},
{
"epoch": 0.6304,
"grad_norm": 0.7345659732818604,
"learning_rate": 1.8107872856913293e-06,
"loss": 0.17963820695877075,
"step": 1970
},
{
"epoch": 0.6336,
"grad_norm": 0.7330230474472046,
"learning_rate": 1.7839829747856096e-06,
"loss": 0.18132896423339845,
"step": 1980
},
{
"epoch": 0.6368,
"grad_norm": 0.7300711870193481,
"learning_rate": 1.7572680330951359e-06,
"loss": 0.17989938259124755,
"step": 1990
},
{
"epoch": 0.64,
"grad_norm": 0.6808901429176331,
"learning_rate": 1.7306457950285747e-06,
"loss": 0.18194899559020997,
"step": 2000
},
{
"epoch": 0.6432,
"grad_norm": 0.7305233478546143,
"learning_rate": 1.704119583423848e-06,
"loss": 0.1753689169883728,
"step": 2010
},
{
"epoch": 0.6464,
"grad_norm": 0.6736240983009338,
"learning_rate": 1.677692709133396e-06,
"loss": 0.1841512680053711,
"step": 2020
},
{
"epoch": 0.6496,
"grad_norm": 0.7294182777404785,
"learning_rate": 1.6513684706109311e-06,
"loss": 0.18403513431549073,
"step": 2030
},
{
"epoch": 0.6528,
"grad_norm": 0.7472907900810242,
"learning_rate": 1.6251501534997529e-06,
"loss": 0.181609046459198,
"step": 2040
},
{
"epoch": 0.656,
"grad_norm": 0.6928985714912415,
"learning_rate": 1.5990410302226405e-06,
"loss": 0.17561020851135253,
"step": 2050
},
{
"epoch": 0.6592,
"grad_norm": 0.7173072695732117,
"learning_rate": 1.5730443595734162e-06,
"loss": 0.18147742748260498,
"step": 2060
},
{
"epoch": 0.6624,
"grad_norm": 0.7320525050163269,
"learning_rate": 1.5471633863101982e-06,
"loss": 0.18284125328063966,
"step": 2070
},
{
"epoch": 0.6656,
"grad_norm": 0.7757209539413452,
"learning_rate": 1.521401340750407e-06,
"loss": 0.18450961112976075,
"step": 2080
},
{
"epoch": 0.6688,
"grad_norm": 0.7066748142242432,
"learning_rate": 1.495761438367577e-06,
"loss": 0.18320014476776122,
"step": 2090
},
{
"epoch": 0.672,
"grad_norm": 0.7570253014564514,
"learning_rate": 1.4702468793900187e-06,
"loss": 0.18169447183609008,
"step": 2100
},
{
"epoch": 0.6752,
"grad_norm": 0.7254899144172668,
"learning_rate": 1.444860848401384e-06,
"loss": 0.1820533037185669,
"step": 2110
},
{
"epoch": 0.6784,
"grad_norm": 0.71804279088974,
"learning_rate": 1.4196065139431866e-06,
"loss": 0.18047101497650148,
"step": 2120
},
{
"epoch": 0.6816,
"grad_norm": 0.7074591517448425,
"learning_rate": 1.3944870281193178e-06,
"loss": 0.18028099536895753,
"step": 2130
},
{
"epoch": 0.6848,
"grad_norm": 0.7010617852210999,
"learning_rate": 1.3695055262026208e-06,
"loss": 0.18579988479614257,
"step": 2140
},
{
"epoch": 0.688,
"grad_norm": 0.6825765371322632,
"learning_rate": 1.3446651262435679e-06,
"loss": 0.17880038022994996,
"step": 2150
},
{
"epoch": 0.6912,
"grad_norm": 0.7335405945777893,
"learning_rate": 1.3199689286810746e-06,
"loss": 0.1810365080833435,
"step": 2160
},
{
"epoch": 0.6944,
"grad_norm": 0.7248731851577759,
"learning_rate": 1.2954200159555294e-06,
"loss": 0.17573955059051513,
"step": 2170
},
{
"epoch": 0.6976,
"grad_norm": 0.7687682509422302,
"learning_rate": 1.2710214521240527e-06,
"loss": 0.1800560474395752,
"step": 2180
},
{
"epoch": 0.7008,
"grad_norm": 0.7647348046302795,
"learning_rate": 1.246776282478063e-06,
"loss": 0.18106215000152587,
"step": 2190
},
{
"epoch": 0.704,
"grad_norm": 0.6903149485588074,
"learning_rate": 1.222687533163181e-06,
"loss": 0.1801396608352661,
"step": 2200
},
{
"epoch": 0.7072,
"grad_norm": 0.8023920655250549,
"learning_rate": 1.1987582108015228e-06,
"loss": 0.18119274377822875,
"step": 2210
},
{
"epoch": 0.7104,
"grad_norm": 0.8501482009887695,
"learning_rate": 1.1749913021164255e-06,
"loss": 0.18241000175476074,
"step": 2220
},
{
"epoch": 0.7136,
"grad_norm": 0.7184765338897705,
"learning_rate": 1.1513897735596702e-06,
"loss": 0.17752939462661743,
"step": 2230
},
{
"epoch": 0.7168,
"grad_norm": 0.7270201444625854,
"learning_rate": 1.127956570941218e-06,
"loss": 0.1741568326950073,
"step": 2240
},
{
"epoch": 0.72,
"grad_norm": 0.8636892437934875,
"learning_rate": 1.104694619061533e-06,
"loss": 0.1828829050064087,
"step": 2250
},
{
"epoch": 0.7232,
"grad_norm": 0.7514937520027161,
"learning_rate": 1.0816068213465295e-06,
"loss": 0.1855594277381897,
"step": 2260
},
{
"epoch": 0.7264,
"grad_norm": 0.7141839265823364,
"learning_rate": 1.0586960594851762e-06,
"loss": 0.18119051456451415,
"step": 2270
},
{
"epoch": 0.7296,
"grad_norm": 0.7850434184074402,
"learning_rate": 1.0359651930698217e-06,
"loss": 0.1793533205986023,
"step": 2280
},
{
"epoch": 0.7328,
"grad_norm": 0.7188661098480225,
"learning_rate": 1.0134170592392837e-06,
"loss": 0.18033140897750854,
"step": 2290
},
{
"epoch": 0.736,
"grad_norm": 0.7670432329177856,
"learning_rate": 9.910544723247204e-07,
"loss": 0.17953202724456788,
"step": 2300
},
{
"epoch": 0.7392,
"grad_norm": 0.7422866225242615,
"learning_rate": 9.688802234983706e-07,
"loss": 0.17864506244659423,
"step": 2310
},
{
"epoch": 0.7424,
"grad_norm": 0.8212261199951172,
"learning_rate": 9.468970804251742e-07,
"loss": 0.18146371841430664,
"step": 2320
},
{
"epoch": 0.7456,
"grad_norm": 0.7881020903587341,
"learning_rate": 9.251077869173244e-07,
"loss": 0.17564767599105835,
"step": 2330
},
{
"epoch": 0.7488,
"grad_norm": 0.7393279671669006,
"learning_rate": 9.035150625918054e-07,
"loss": 0.1761918306350708,
"step": 2340
},
{
"epoch": 0.752,
"grad_norm": 0.7876865863800049,
"learning_rate": 8.821216025309395e-07,
"loss": 0.17920515537261963,
"step": 2350
},
{
"epoch": 0.7552,
"grad_norm": 0.7123493552207947,
"learning_rate": 8.609300769460055e-07,
"loss": 0.17455912828445436,
"step": 2360
},
{
"epoch": 0.7584,
"grad_norm": 0.7895896434783936,
"learning_rate": 8.399431308439592e-07,
"loss": 0.1802891969680786,
"step": 2370
},
{
"epoch": 0.7616,
"grad_norm": 0.7665727138519287,
"learning_rate": 8.191633836972962e-07,
"loss": 0.18085892200469972,
"step": 2380
},
{
"epoch": 0.7648,
"grad_norm": 0.7168281674385071,
"learning_rate": 7.985934291171024e-07,
"loss": 0.1772964835166931,
"step": 2390
},
{
"epoch": 0.768,
"grad_norm": 0.7637878656387329,
"learning_rate": 7.7823583452934e-07,
"loss": 0.18090858459472656,
"step": 2400
},
{
"epoch": 0.7712,
"grad_norm": 0.7479363679885864,
"learning_rate": 7.58093140854389e-07,
"loss": 0.1801297664642334,
"step": 2410
},
{
"epoch": 0.7744,
"grad_norm": 0.7759881615638733,
"learning_rate": 7.381678621899077e-07,
"loss": 0.18061695098876954,
"step": 2420
},
{
"epoch": 0.7776,
"grad_norm": 0.7594044804573059,
"learning_rate": 7.184624854970379e-07,
"loss": 0.17679831981658936,
"step": 2430
},
{
"epoch": 0.7808,
"grad_norm": 0.6967716813087463,
"learning_rate": 6.989794702899932e-07,
"loss": 0.17618390321731567,
"step": 2440
},
{
"epoch": 0.784,
"grad_norm": 0.7475560307502747,
"learning_rate": 6.797212483290777e-07,
"loss": 0.1785645604133606,
"step": 2450
},
{
"epoch": 0.7872,
"grad_norm": 0.7539611458778381,
"learning_rate": 6.60690223317171e-07,
"loss": 0.17528293132781983,
"step": 2460
},
{
"epoch": 0.7904,
"grad_norm": 0.7217035293579102,
"learning_rate": 6.418887705997046e-07,
"loss": 0.17803908586502076,
"step": 2470
},
{
"epoch": 0.7936,
"grad_norm": 0.9057507514953613,
"learning_rate": 6.23319236868189e-07,
"loss": 0.1803913116455078,
"step": 2480
},
{
"epoch": 0.7968,
"grad_norm": 0.7403589487075806,
"learning_rate": 6.049839398673141e-07,
"loss": 0.18166184425354004,
"step": 2490
},
{
"epoch": 0.8,
"grad_norm": 0.7768926620483398,
"learning_rate": 5.868851681056567e-07,
"loss": 0.18331187963485718,
"step": 2500
},
{
"epoch": 0.8032,
"grad_norm": 0.7482465505599976,
"learning_rate": 5.690251805700467e-07,
"loss": 0.18083620071411133,
"step": 2510
},
{
"epoch": 0.8064,
"grad_norm": 0.8199617266654968,
"learning_rate": 5.514062064436096e-07,
"loss": 0.1817050337791443,
"step": 2520
},
{
"epoch": 0.8096,
"grad_norm": 0.8081801533699036,
"learning_rate": 5.34030444827533e-07,
"loss": 0.17815799713134767,
"step": 2530
},
{
"epoch": 0.8128,
"grad_norm": 0.7771419286727905,
"learning_rate": 5.169000644665895e-07,
"loss": 0.17568042278289794,
"step": 2540
},
{
"epoch": 0.816,
"grad_norm": 0.7164714336395264,
"learning_rate": 5.000172034784442e-07,
"loss": 0.1774258255958557,
"step": 2550
},
{
"epoch": 0.8192,
"grad_norm": 0.8132921457290649,
"learning_rate": 4.833839690867853e-07,
"loss": 0.1790420413017273,
"step": 2560
},
{
"epoch": 0.8224,
"grad_norm": 0.7847883105278015,
"learning_rate": 4.6700243735831705e-07,
"loss": 0.17356895208358764,
"step": 2570
},
{
"epoch": 0.8256,
"grad_norm": 0.7992151379585266,
"learning_rate": 4.508746529436311e-07,
"loss": 0.1719137191772461,
"step": 2580
},
{
"epoch": 0.8288,
"grad_norm": 0.7740475535392761,
"learning_rate": 4.350026288220083e-07,
"loss": 0.17878987789154052,
"step": 2590
},
{
"epoch": 0.832,
"grad_norm": 0.751238226890564,
"learning_rate": 4.1938834605017133e-07,
"loss": 0.17630958557128906,
"step": 2600
},
{
"epoch": 0.8352,
"grad_norm": 0.7768763303756714,
"learning_rate": 4.0403375351501515e-07,
"loss": 0.17546494007110597,
"step": 2610
},
{
"epoch": 0.8384,
"grad_norm": 0.7661908864974976,
"learning_rate": 3.88940767690362e-07,
"loss": 0.17503058910369873,
"step": 2620
},
{
"epoch": 0.8416,
"grad_norm": 0.8674443960189819,
"learning_rate": 3.7411127239775774e-07,
"loss": 0.1738388180732727,
"step": 2630
},
{
"epoch": 0.8448,
"grad_norm": 0.8360145688056946,
"learning_rate": 3.595471185713431e-07,
"loss": 0.1726490616798401,
"step": 2640
},
{
"epoch": 0.848,
"grad_norm": 0.7894387245178223,
"learning_rate": 3.4525012402682826e-07,
"loss": 0.17366337776184082,
"step": 2650
},
{
"epoch": 0.8512,
"grad_norm": 0.793187141418457,
"learning_rate": 3.3122207323460804e-07,
"loss": 0.17671537399291992,
"step": 2660
},
{
"epoch": 0.8544,
"grad_norm": 0.7538631558418274,
"learning_rate": 3.1746471709702963e-07,
"loss": 0.1724637746810913,
"step": 2670
},
{
"epoch": 0.8576,
"grad_norm": 0.7905218005180359,
"learning_rate": 3.039797727298585e-07,
"loss": 0.17893683910369873,
"step": 2680
},
{
"epoch": 0.8608,
"grad_norm": 0.8051464557647705,
"learning_rate": 2.9076892324795546e-07,
"loss": 0.17726495265960693,
"step": 2690
},
{
"epoch": 0.864,
"grad_norm": 0.7938066720962524,
"learning_rate": 2.778338175551995e-07,
"loss": 0.1730543851852417,
"step": 2700
},
{
"epoch": 0.8672,
"grad_norm": 0.8020470142364502,
"learning_rate": 2.6517607013868326e-07,
"loss": 0.1814263343811035,
"step": 2710
},
{
"epoch": 0.8704,
"grad_norm": 0.8322368264198303,
"learning_rate": 2.527972608672002e-07,
"loss": 0.17693625688552855,
"step": 2720
},
{
"epoch": 0.8736,
"grad_norm": 0.8700771331787109,
"learning_rate": 2.40698934794053e-07,
"loss": 0.17613067626953124,
"step": 2730
},
{
"epoch": 0.8768,
"grad_norm": 0.8105334043502808,
"learning_rate": 2.2888260196421237e-07,
"loss": 0.175561785697937,
"step": 2740
},
{
"epoch": 0.88,
"grad_norm": 0.7427732348442078,
"learning_rate": 2.1734973722583735e-07,
"loss": 0.1789578080177307,
"step": 2750
},
{
"epoch": 0.8832,
"grad_norm": 0.7436424493789673,
"learning_rate": 2.0610178004619564e-07,
"loss": 0.17029953002929688,
"step": 2760
},
{
"epoch": 0.8864,
"grad_norm": 0.7291203737258911,
"learning_rate": 1.9514013433199834e-07,
"loss": 0.18220956325531007,
"step": 2770
},
{
"epoch": 0.8896,
"grad_norm": 0.7028378248214722,
"learning_rate": 1.8446616825416958e-07,
"loss": 0.18021023273468018,
"step": 2780
},
{
"epoch": 0.8928,
"grad_norm": 0.8554368615150452,
"learning_rate": 1.7408121407708007e-07,
"loss": 0.18080689907073974,
"step": 2790
},
{
"epoch": 0.896,
"grad_norm": 0.8060093522071838,
"learning_rate": 1.6398656799226253e-07,
"loss": 0.16961441040039063,
"step": 2800
},
{
"epoch": 0.8992,
"grad_norm": 0.7748635411262512,
"learning_rate": 1.5418348995662773e-07,
"loss": 0.17575571537017823,
"step": 2810
},
{
"epoch": 0.9024,
"grad_norm": 0.8475610613822937,
"learning_rate": 1.4467320353520275e-07,
"loss": 0.17327470779418946,
"step": 2820
},
{
"epoch": 0.9056,
"grad_norm": 0.8207055926322937,
"learning_rate": 1.3545689574841341e-07,
"loss": 0.17445048093795776,
"step": 2830
},
{
"epoch": 0.9088,
"grad_norm": 0.8258077502250671,
"learning_rate": 1.26535716923927e-07,
"loss": 0.18045923709869385,
"step": 2840
},
{
"epoch": 0.912,
"grad_norm": 0.7929080724716187,
"learning_rate": 1.1791078055307493e-07,
"loss": 0.17697184085845946,
"step": 2850
},
{
"epoch": 0.9152,
"grad_norm": 0.8584709167480469,
"learning_rate": 1.0958316315187289e-07,
"loss": 0.1764688014984131,
"step": 2860
},
{
"epoch": 0.9184,
"grad_norm": 0.7925869822502136,
"learning_rate": 1.0155390412665528e-07,
"loss": 0.17493282556533812,
"step": 2870
},
{
"epoch": 0.9216,
"grad_norm": 0.7998941540718079,
"learning_rate": 9.38240056443443e-08,
"loss": 0.17196444272994996,
"step": 2880
},
{
"epoch": 0.9248,
"grad_norm": 0.771435022354126,
"learning_rate": 8.639443250736402e-08,
"loss": 0.17460463047027588,
"step": 2890
},
{
"epoch": 0.928,
"grad_norm": 0.7990598082542419,
"learning_rate": 7.926611203321777e-08,
"loss": 0.17567555904388427,
"step": 2900
},
{
"epoch": 0.9312,
"grad_norm": 0.8290013670921326,
"learning_rate": 7.243993393874882e-08,
"loss": 0.17620582580566407,
"step": 2910
},
{
"epoch": 0.9344,
"grad_norm": 0.7977768778800964,
"learning_rate": 6.591675022908805e-08,
"loss": 0.17442046403884887,
"step": 2920
},
{
"epoch": 0.9376,
"grad_norm": 0.8192407488822937,
"learning_rate": 5.969737509131241e-08,
"loss": 0.17177685499191284,
"step": 2930
},
{
"epoch": 0.9408,
"grad_norm": 0.7740393877029419,
"learning_rate": 5.3782584792823334e-08,
"loss": 0.1760073184967041,
"step": 2940
},
{
"epoch": 0.944,
"grad_norm": 0.7851013541221619,
"learning_rate": 4.817311758445686e-08,
"loss": 0.1770064949989319,
"step": 2950
},
{
"epoch": 0.9472,
"grad_norm": 0.7475484609603882,
"learning_rate": 4.286967360833866e-08,
"loss": 0.1779289126396179,
"step": 2960
},
{
"epoch": 0.9504,
"grad_norm": 0.7964699268341064,
"learning_rate": 3.787291481049754e-08,
"loss": 0.17799465656280516,
"step": 2970
},
{
"epoch": 0.9536,
"grad_norm": 0.8085147738456726,
"learning_rate": 3.3183464858244364e-08,
"loss": 0.18326969146728517,
"step": 2980
},
{
"epoch": 0.9568,
"grad_norm": 0.8301185369491577,
"learning_rate": 2.8801909062328992e-08,
"loss": 0.16990091800689697,
"step": 2990
},
{
"epoch": 0.96,
"grad_norm": 0.7779503464698792,
"learning_rate": 2.4728794303886248e-08,
"loss": 0.1689441442489624,
"step": 3000
},
{
"epoch": 0.9632,
"grad_norm": 0.8286770582199097,
"learning_rate": 2.0964628966175794e-08,
"loss": 0.17429101467132568,
"step": 3010
},
{
"epoch": 0.9664,
"grad_norm": 0.7744504809379578,
"learning_rate": 1.750988287113009e-08,
"loss": 0.17314184904098512,
"step": 3020
},
{
"epoch": 0.9696,
"grad_norm": 0.8272920846939087,
"learning_rate": 1.4364987220713278e-08,
"loss": 0.18057892322540284,
"step": 3030
},
{
"epoch": 0.9728,
"grad_norm": 0.7745753526687622,
"learning_rate": 1.1530334543099763e-08,
"loss": 0.17820963859558106,
"step": 3040
},
{
"epoch": 0.976,
"grad_norm": 0.8267188668251038,
"learning_rate": 9.006278643683697e-09,
"loss": 0.1798885703086853,
"step": 3050
},
{
"epoch": 0.9792,
"grad_norm": 0.7813923954963684,
"learning_rate": 6.793134560916514e-09,
"loss": 0.18165643215179444,
"step": 3060
},
{
"epoch": 0.9824,
"grad_norm": 0.735295832157135,
"learning_rate": 4.891178526986451e-09,
"loss": 0.1776532292366028,
"step": 3070
},
{
"epoch": 0.9856,
"grad_norm": 0.8481670022010803,
"learning_rate": 3.3006479333413943e-09,
"loss": 0.1802857756614685,
"step": 3080
},
{
"epoch": 0.9888,
"grad_norm": 0.7306717038154602,
"learning_rate": 2.021741301058422e-09,
"loss": 0.17649333477020263,
"step": 3090
},
{
"epoch": 0.992,
"grad_norm": 0.8213961124420166,
"learning_rate": 1.0546182560652872e-09,
"loss": 0.1785252332687378,
"step": 3100
},
{
"epoch": 0.9952,
"grad_norm": 0.7591016888618469,
"learning_rate": 3.9939950921774607e-10,
"loss": 0.1741119384765625,
"step": 3110
},
{
"epoch": 0.9984,
"grad_norm": 0.7711865901947021,
"learning_rate": 5.616684123160854e-11,
"loss": 0.17355493307113648,
"step": 3120
},
{
"epoch": 1.0,
"step": 3125,
"total_flos": 4.165948759592665e+18,
"train_loss": 0.19418699533462525,
"train_runtime": 24088.1057,
"train_samples_per_second": 8.303,
"train_steps_per_second": 0.13
}
],
"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": 4.165948759592665e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}