Safetensors
CosFly-Track / checkpoints /Qwen3.5-9B-SFT /trainer_state.json
Ys404's picture
Add scripts and checkpoints (CosFly-Track release)
fd8aeaf verified
raw
history blame
55.5 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": 14.152398109436035,
"learning_rate": 1.437699680511182e-07,
"loss": 0.37259225845336913,
"step": 10
},
{
"epoch": 0.0064,
"grad_norm": 5.804495334625244,
"learning_rate": 3.0351437699680514e-07,
"loss": 0.3420424461364746,
"step": 20
},
{
"epoch": 0.0096,
"grad_norm": 1.7388691902160645,
"learning_rate": 4.6325878594249205e-07,
"loss": 0.2912754058837891,
"step": 30
},
{
"epoch": 0.0128,
"grad_norm": 0.9213385581970215,
"learning_rate": 6.230031948881789e-07,
"loss": 0.2695852518081665,
"step": 40
},
{
"epoch": 0.016,
"grad_norm": 0.7402991056442261,
"learning_rate": 7.82747603833866e-07,
"loss": 0.2490011692047119,
"step": 50
},
{
"epoch": 0.0192,
"grad_norm": 0.8080454468727112,
"learning_rate": 9.424920127795528e-07,
"loss": 0.24377479553222656,
"step": 60
},
{
"epoch": 0.0224,
"grad_norm": 0.7839148640632629,
"learning_rate": 1.1022364217252397e-06,
"loss": 0.2515081882476807,
"step": 70
},
{
"epoch": 0.0256,
"grad_norm": 0.6039342880249023,
"learning_rate": 1.2619808306709266e-06,
"loss": 0.236421799659729,
"step": 80
},
{
"epoch": 0.0288,
"grad_norm": 0.6679620146751404,
"learning_rate": 1.4217252396166134e-06,
"loss": 0.23224935531616211,
"step": 90
},
{
"epoch": 0.032,
"grad_norm": 0.6516961455345154,
"learning_rate": 1.5814696485623005e-06,
"loss": 0.23475091457366942,
"step": 100
},
{
"epoch": 0.0352,
"grad_norm": 0.6763356924057007,
"learning_rate": 1.7412140575079875e-06,
"loss": 0.2348011016845703,
"step": 110
},
{
"epoch": 0.0384,
"grad_norm": 0.7100752592086792,
"learning_rate": 1.9009584664536742e-06,
"loss": 0.22568964958190918,
"step": 120
},
{
"epoch": 0.0416,
"grad_norm": 0.7218499779701233,
"learning_rate": 2.060702875399361e-06,
"loss": 0.22881340980529785,
"step": 130
},
{
"epoch": 0.0448,
"grad_norm": 0.7274235486984253,
"learning_rate": 2.220447284345048e-06,
"loss": 0.2210618019104004,
"step": 140
},
{
"epoch": 0.048,
"grad_norm": 0.7489368319511414,
"learning_rate": 2.380191693290735e-06,
"loss": 0.21785550117492675,
"step": 150
},
{
"epoch": 0.0512,
"grad_norm": 0.636695384979248,
"learning_rate": 2.539936102236422e-06,
"loss": 0.21415135860443116,
"step": 160
},
{
"epoch": 0.0544,
"grad_norm": 0.7647548317909241,
"learning_rate": 2.699680511182109e-06,
"loss": 0.22171540260314943,
"step": 170
},
{
"epoch": 0.0576,
"grad_norm": 0.6201676726341248,
"learning_rate": 2.8594249201277955e-06,
"loss": 0.21528491973876954,
"step": 180
},
{
"epoch": 0.0608,
"grad_norm": 0.7164552807807922,
"learning_rate": 3.0191693290734825e-06,
"loss": 0.21876811981201172,
"step": 190
},
{
"epoch": 0.064,
"grad_norm": 0.7416535019874573,
"learning_rate": 3.17891373801917e-06,
"loss": 0.22004494667053223,
"step": 200
},
{
"epoch": 0.0672,
"grad_norm": 0.6418226361274719,
"learning_rate": 3.3386581469648564e-06,
"loss": 0.21785356998443603,
"step": 210
},
{
"epoch": 0.0704,
"grad_norm": 0.5532397031784058,
"learning_rate": 3.4984025559105434e-06,
"loss": 0.2117140293121338,
"step": 220
},
{
"epoch": 0.0736,
"grad_norm": 0.6566811800003052,
"learning_rate": 3.6581469648562303e-06,
"loss": 0.21361856460571288,
"step": 230
},
{
"epoch": 0.0768,
"grad_norm": 0.5842558741569519,
"learning_rate": 3.817891373801918e-06,
"loss": 0.21009182929992676,
"step": 240
},
{
"epoch": 0.08,
"grad_norm": 0.5284646153450012,
"learning_rate": 3.977635782747604e-06,
"loss": 0.20717477798461914,
"step": 250
},
{
"epoch": 0.0832,
"grad_norm": 0.4791855812072754,
"learning_rate": 4.137380191693291e-06,
"loss": 0.20851461887359618,
"step": 260
},
{
"epoch": 0.0864,
"grad_norm": 0.5538341999053955,
"learning_rate": 4.297124600638978e-06,
"loss": 0.21010227203369142,
"step": 270
},
{
"epoch": 0.0896,
"grad_norm": 0.5214399695396423,
"learning_rate": 4.456869009584665e-06,
"loss": 0.20617682933807374,
"step": 280
},
{
"epoch": 0.0928,
"grad_norm": 0.5693404674530029,
"learning_rate": 4.616613418530352e-06,
"loss": 0.21186504364013672,
"step": 290
},
{
"epoch": 0.096,
"grad_norm": 0.520547091960907,
"learning_rate": 4.776357827476039e-06,
"loss": 0.2132932186126709,
"step": 300
},
{
"epoch": 0.0992,
"grad_norm": 0.5054877400398254,
"learning_rate": 4.936102236421725e-06,
"loss": 0.2020902156829834,
"step": 310
},
{
"epoch": 0.1024,
"grad_norm": 0.5992368459701538,
"learning_rate": 4.999943833158769e-06,
"loss": 0.20265860557556153,
"step": 320
},
{
"epoch": 0.1056,
"grad_norm": 0.4485788941383362,
"learning_rate": 4.999600600490783e-06,
"loss": 0.2094430923461914,
"step": 330
},
{
"epoch": 0.1088,
"grad_norm": 0.48217350244522095,
"learning_rate": 4.9989453817439345e-06,
"loss": 0.20311150550842286,
"step": 340
},
{
"epoch": 0.112,
"grad_norm": 0.48118671774864197,
"learning_rate": 4.997978258698942e-06,
"loss": 0.20234436988830568,
"step": 350
},
{
"epoch": 0.1152,
"grad_norm": 0.4691317677497864,
"learning_rate": 4.996699352066659e-06,
"loss": 0.20289850234985352,
"step": 360
},
{
"epoch": 0.1184,
"grad_norm": 0.5056607723236084,
"learning_rate": 4.995108821473014e-06,
"loss": 0.20348825454711914,
"step": 370
},
{
"epoch": 0.1216,
"grad_norm": 0.46245047450065613,
"learning_rate": 4.993206865439084e-06,
"loss": 0.1985713481903076,
"step": 380
},
{
"epoch": 0.1248,
"grad_norm": 0.45679083466529846,
"learning_rate": 4.990993721356317e-06,
"loss": 0.1985337495803833,
"step": 390
},
{
"epoch": 0.128,
"grad_norm": 0.513569712638855,
"learning_rate": 4.988469665456901e-06,
"loss": 0.1989649534225464,
"step": 400
},
{
"epoch": 0.1312,
"grad_norm": 0.47685959935188293,
"learning_rate": 4.985635012779288e-06,
"loss": 0.20632779598236084,
"step": 410
},
{
"epoch": 0.1344,
"grad_norm": 0.42005518078804016,
"learning_rate": 4.98249011712887e-06,
"loss": 0.1997704029083252,
"step": 420
},
{
"epoch": 0.1376,
"grad_norm": 0.40938711166381836,
"learning_rate": 4.979035371033824e-06,
"loss": 0.198714542388916,
"step": 430
},
{
"epoch": 0.1408,
"grad_norm": 0.4478866755962372,
"learning_rate": 4.975271205696115e-06,
"loss": 0.1973956823348999,
"step": 440
},
{
"epoch": 0.144,
"grad_norm": 0.4451947510242462,
"learning_rate": 4.971198090937671e-06,
"loss": 0.19358736276626587,
"step": 450
},
{
"epoch": 0.1472,
"grad_norm": 0.4245293140411377,
"learning_rate": 4.966816535141756e-06,
"loss": 0.18962936401367186,
"step": 460
},
{
"epoch": 0.1504,
"grad_norm": 0.44330471754074097,
"learning_rate": 4.9621270851895035e-06,
"loss": 0.19763081073760985,
"step": 470
},
{
"epoch": 0.1536,
"grad_norm": 0.4569932520389557,
"learning_rate": 4.957130326391662e-06,
"loss": 0.19951647520065308,
"step": 480
},
{
"epoch": 0.1568,
"grad_norm": 0.46161818504333496,
"learning_rate": 4.951826882415544e-06,
"loss": 0.1966404676437378,
"step": 490
},
{
"epoch": 0.16,
"grad_norm": 0.5257421135902405,
"learning_rate": 4.946217415207177e-06,
"loss": 0.1917549967765808,
"step": 500
},
{
"epoch": 0.1632,
"grad_norm": 0.47060427069664,
"learning_rate": 4.940302624908689e-06,
"loss": 0.1965235710144043,
"step": 510
},
{
"epoch": 0.1664,
"grad_norm": 0.49784040451049805,
"learning_rate": 4.934083249770912e-06,
"loss": 0.19047110080718993,
"step": 520
},
{
"epoch": 0.1696,
"grad_norm": 0.46195048093795776,
"learning_rate": 4.927560066061251e-06,
"loss": 0.1933290719985962,
"step": 530
},
{
"epoch": 0.1728,
"grad_norm": 0.44262751936912537,
"learning_rate": 4.920733887966783e-06,
"loss": 0.20422852039337158,
"step": 540
},
{
"epoch": 0.176,
"grad_norm": 0.4286530315876007,
"learning_rate": 4.913605567492636e-06,
"loss": 0.19317498207092285,
"step": 550
},
{
"epoch": 0.1792,
"grad_norm": 0.4754306375980377,
"learning_rate": 4.906175994355656e-06,
"loss": 0.19484668970108032,
"step": 560
},
{
"epoch": 0.1824,
"grad_norm": 0.542935311794281,
"learning_rate": 4.898446095873345e-06,
"loss": 0.19600849151611327,
"step": 570
},
{
"epoch": 0.1856,
"grad_norm": 0.4545859396457672,
"learning_rate": 4.890416836848128e-06,
"loss": 0.193756628036499,
"step": 580
},
{
"epoch": 0.1888,
"grad_norm": 0.44452813267707825,
"learning_rate": 4.882089219446925e-06,
"loss": 0.18871793746948243,
"step": 590
},
{
"epoch": 0.192,
"grad_norm": 0.49089452624320984,
"learning_rate": 4.873464283076074e-06,
"loss": 0.19349279403686523,
"step": 600
},
{
"epoch": 0.1952,
"grad_norm": 0.40194588899612427,
"learning_rate": 4.864543104251587e-06,
"loss": 0.19310340881347657,
"step": 610
},
{
"epoch": 0.1984,
"grad_norm": 0.4422169327735901,
"learning_rate": 4.855326796464798e-06,
"loss": 0.1940554141998291,
"step": 620
},
{
"epoch": 0.2016,
"grad_norm": 0.3977762460708618,
"learning_rate": 4.8458165100433725e-06,
"loss": 0.1905130386352539,
"step": 630
},
{
"epoch": 0.2048,
"grad_norm": 0.47129112482070923,
"learning_rate": 4.836013432007738e-06,
"loss": 0.1905440092086792,
"step": 640
},
{
"epoch": 0.208,
"grad_norm": 0.44452038407325745,
"learning_rate": 4.825918785922921e-06,
"loss": 0.19458037614822388,
"step": 650
},
{
"epoch": 0.2112,
"grad_norm": 0.511098325252533,
"learning_rate": 4.8155338317458315e-06,
"loss": 0.19132745265960693,
"step": 660
},
{
"epoch": 0.2144,
"grad_norm": 0.44879648089408875,
"learning_rate": 4.804859865668002e-06,
"loss": 0.18898229598999022,
"step": 670
},
{
"epoch": 0.2176,
"grad_norm": 0.42120668292045593,
"learning_rate": 4.793898219953804e-06,
"loss": 0.1905623197555542,
"step": 680
},
{
"epoch": 0.2208,
"grad_norm": 0.4307490587234497,
"learning_rate": 4.782650262774164e-06,
"loss": 0.1913731336593628,
"step": 690
},
{
"epoch": 0.224,
"grad_norm": 0.4430791735649109,
"learning_rate": 4.7711173980357886e-06,
"loss": 0.1875431776046753,
"step": 700
},
{
"epoch": 0.2272,
"grad_norm": 0.45506489276885986,
"learning_rate": 4.759301065205947e-06,
"loss": 0.18780747652053834,
"step": 710
},
{
"epoch": 0.2304,
"grad_norm": 0.4784105122089386,
"learning_rate": 4.7472027391328e-06,
"loss": 0.18949007987976074,
"step": 720
},
{
"epoch": 0.2336,
"grad_norm": 0.5166275501251221,
"learning_rate": 4.734823929861317e-06,
"loss": 0.19141184091567992,
"step": 730
},
{
"epoch": 0.2368,
"grad_norm": 0.4763261675834656,
"learning_rate": 4.722166182444801e-06,
"loss": 0.1886807918548584,
"step": 740
},
{
"epoch": 0.24,
"grad_norm": 0.4896913170814514,
"learning_rate": 4.709231076752045e-06,
"loss": 0.18765724897384645,
"step": 750
},
{
"epoch": 0.2432,
"grad_norm": 0.4405442178249359,
"learning_rate": 4.696020227270142e-06,
"loss": 0.1914137840270996,
"step": 760
},
{
"epoch": 0.2464,
"grad_norm": 0.4831804633140564,
"learning_rate": 4.6825352829029705e-06,
"loss": 0.1876143217086792,
"step": 770
},
{
"epoch": 0.2496,
"grad_norm": 0.541253387928009,
"learning_rate": 4.668777926765392e-06,
"loss": 0.18711633682250978,
"step": 780
},
{
"epoch": 0.2528,
"grad_norm": 0.4409460723400116,
"learning_rate": 4.6547498759731725e-06,
"loss": 0.182257080078125,
"step": 790
},
{
"epoch": 0.256,
"grad_norm": 0.4363306164741516,
"learning_rate": 4.6404528814286575e-06,
"loss": 0.18522121906280517,
"step": 800
},
{
"epoch": 0.2592,
"grad_norm": 0.42511361837387085,
"learning_rate": 4.6258887276022425e-06,
"loss": 0.19108107089996337,
"step": 810
},
{
"epoch": 0.2624,
"grad_norm": 0.549828290939331,
"learning_rate": 4.611059232309639e-06,
"loss": 0.18742183446884156,
"step": 820
},
{
"epoch": 0.2656,
"grad_norm": 0.48544079065322876,
"learning_rate": 4.595966246484986e-06,
"loss": 0.18868749141693114,
"step": 830
},
{
"epoch": 0.2688,
"grad_norm": 0.4877803325653076,
"learning_rate": 4.580611653949829e-06,
"loss": 0.1887261152267456,
"step": 840
},
{
"epoch": 0.272,
"grad_norm": 0.43121787905693054,
"learning_rate": 4.564997371177992e-06,
"loss": 0.18640785217285155,
"step": 850
},
{
"epoch": 0.2752,
"grad_norm": 0.46837401390075684,
"learning_rate": 4.54912534705637e-06,
"loss": 0.1863306999206543,
"step": 860
},
{
"epoch": 0.2784,
"grad_norm": 0.4588525593280792,
"learning_rate": 4.532997562641683e-06,
"loss": 0.18255772590637206,
"step": 870
},
{
"epoch": 0.2816,
"grad_norm": 0.4227045774459839,
"learning_rate": 4.516616030913214e-06,
"loss": 0.18717021942138673,
"step": 880
},
{
"epoch": 0.2848,
"grad_norm": 0.4720568358898163,
"learning_rate": 4.499982796521556e-06,
"loss": 0.18217651844024657,
"step": 890
},
{
"epoch": 0.288,
"grad_norm": 0.4398342967033386,
"learning_rate": 4.48309993553341e-06,
"loss": 0.18433446884155275,
"step": 900
},
{
"epoch": 0.2912,
"grad_norm": 0.44484105706214905,
"learning_rate": 4.465969555172468e-06,
"loss": 0.18349353075027466,
"step": 910
},
{
"epoch": 0.2944,
"grad_norm": 0.43579065799713135,
"learning_rate": 4.448593793556391e-06,
"loss": 0.1893969178199768,
"step": 920
},
{
"epoch": 0.2976,
"grad_norm": 0.43624043464660645,
"learning_rate": 4.430974819429954e-06,
"loss": 0.18258814811706542,
"step": 930
},
{
"epoch": 0.3008,
"grad_norm": 0.42816856503486633,
"learning_rate": 4.413114831894344e-06,
"loss": 0.1781400442123413,
"step": 940
},
{
"epoch": 0.304,
"grad_norm": 0.4396454989910126,
"learning_rate": 4.3950160601326865e-06,
"loss": 0.17804189920425414,
"step": 950
},
{
"epoch": 0.3072,
"grad_norm": 19.137165069580078,
"learning_rate": 4.376680763131811e-06,
"loss": 0.18540620803833008,
"step": 960
},
{
"epoch": 0.3104,
"grad_norm": 0.4217820167541504,
"learning_rate": 4.358111229400296e-06,
"loss": 0.19113874435424805,
"step": 970
},
{
"epoch": 0.3136,
"grad_norm": 0.44505834579467773,
"learning_rate": 4.33930977668283e-06,
"loss": 0.18937530517578124,
"step": 980
},
{
"epoch": 0.3168,
"grad_norm": 0.5085921287536621,
"learning_rate": 4.320278751670922e-06,
"loss": 0.18254513740539552,
"step": 990
},
{
"epoch": 0.32,
"grad_norm": 0.49309051036834717,
"learning_rate": 4.301020529710009e-06,
"loss": 0.18421100378036498,
"step": 1000
},
{
"epoch": 0.3232,
"grad_norm": 0.47118672728538513,
"learning_rate": 4.281537514502962e-06,
"loss": 0.1877603769302368,
"step": 1010
},
{
"epoch": 0.3264,
"grad_norm": 0.4600847363471985,
"learning_rate": 4.261832137810093e-06,
"loss": 0.1849065065383911,
"step": 1020
},
{
"epoch": 0.3296,
"grad_norm": 0.44828227162361145,
"learning_rate": 4.241906859145611e-06,
"loss": 0.18526692390441896,
"step": 1030
},
{
"epoch": 0.3328,
"grad_norm": 0.4324962794780731,
"learning_rate": 4.221764165470661e-06,
"loss": 0.18226839303970338,
"step": 1040
},
{
"epoch": 0.336,
"grad_norm": 0.5017353296279907,
"learning_rate": 4.201406570882898e-06,
"loss": 0.181982159614563,
"step": 1050
},
{
"epoch": 0.3392,
"grad_norm": 0.44924962520599365,
"learning_rate": 4.180836616302704e-06,
"loss": 0.18122615814208984,
"step": 1060
},
{
"epoch": 0.3424,
"grad_norm": 0.5138829350471497,
"learning_rate": 4.160056869156041e-06,
"loss": 0.18444604873657228,
"step": 1070
},
{
"epoch": 0.3456,
"grad_norm": 0.41282355785369873,
"learning_rate": 4.139069923053995e-06,
"loss": 0.18227832317352294,
"step": 1080
},
{
"epoch": 0.3488,
"grad_norm": 0.4269866943359375,
"learning_rate": 4.117878397469062e-06,
"loss": 0.18615881204605103,
"step": 1090
},
{
"epoch": 0.352,
"grad_norm": 0.39903032779693604,
"learning_rate": 4.096484937408195e-06,
"loss": 0.17755311727523804,
"step": 1100
},
{
"epoch": 0.3552,
"grad_norm": 0.4636191725730896,
"learning_rate": 4.074892213082676e-06,
"loss": 0.17751554250717164,
"step": 1110
},
{
"epoch": 0.3584,
"grad_norm": 0.48484447598457336,
"learning_rate": 4.0531029195748265e-06,
"loss": 0.18447484970092773,
"step": 1120
},
{
"epoch": 0.3616,
"grad_norm": 0.4330807030200958,
"learning_rate": 4.03111977650163e-06,
"loss": 0.18296458721160888,
"step": 1130
},
{
"epoch": 0.3648,
"grad_norm": 0.4666971266269684,
"learning_rate": 4.008945527675281e-06,
"loss": 0.18310744762420655,
"step": 1140
},
{
"epoch": 0.368,
"grad_norm": 0.46516141295433044,
"learning_rate": 3.986582940760717e-06,
"loss": 0.1737138032913208,
"step": 1150
},
{
"epoch": 0.3712,
"grad_norm": 0.47476157546043396,
"learning_rate": 3.9640348069301785e-06,
"loss": 0.1797748327255249,
"step": 1160
},
{
"epoch": 0.3744,
"grad_norm": 0.4147510230541229,
"learning_rate": 3.941303940514826e-06,
"loss": 0.18082959651947023,
"step": 1170
},
{
"epoch": 0.3776,
"grad_norm": 0.4981541633605957,
"learning_rate": 3.918393178653472e-06,
"loss": 0.1821175694465637,
"step": 1180
},
{
"epoch": 0.3808,
"grad_norm": 0.44226428866386414,
"learning_rate": 3.895305380938468e-06,
"loss": 0.1790626049041748,
"step": 1190
},
{
"epoch": 0.384,
"grad_norm": 0.44139742851257324,
"learning_rate": 3.872043429058783e-06,
"loss": 0.1771311044692993,
"step": 1200
},
{
"epoch": 0.3872,
"grad_norm": 0.4409257173538208,
"learning_rate": 3.84861022644033e-06,
"loss": 0.17819294929504395,
"step": 1210
},
{
"epoch": 0.3904,
"grad_norm": 0.4447258412837982,
"learning_rate": 3.825008697883574e-06,
"loss": 0.1871392011642456,
"step": 1220
},
{
"epoch": 0.3936,
"grad_norm": 0.4376235008239746,
"learning_rate": 3.8012417891984776e-06,
"loss": 0.1800956606864929,
"step": 1230
},
{
"epoch": 0.3968,
"grad_norm": 0.484053373336792,
"learning_rate": 3.777312466836819e-06,
"loss": 0.17867467403411866,
"step": 1240
},
{
"epoch": 0.4,
"grad_norm": 0.4140814542770386,
"learning_rate": 3.7532237175219378e-06,
"loss": 0.17768914699554444,
"step": 1250
},
{
"epoch": 0.4032,
"grad_norm": 0.46091777086257935,
"learning_rate": 3.728978547875948e-06,
"loss": 0.1821272373199463,
"step": 1260
},
{
"epoch": 0.4064,
"grad_norm": 0.48253530263900757,
"learning_rate": 3.7045799840444712e-06,
"loss": 0.1769007444381714,
"step": 1270
},
{
"epoch": 0.4096,
"grad_norm": 0.4717654287815094,
"learning_rate": 3.6800310713189258e-06,
"loss": 0.17760204076766967,
"step": 1280
},
{
"epoch": 0.4128,
"grad_norm": 0.510908305644989,
"learning_rate": 3.6553348737564328e-06,
"loss": 0.17857449054718016,
"step": 1290
},
{
"epoch": 0.416,
"grad_norm": 0.41923636198043823,
"learning_rate": 3.6304944737973794e-06,
"loss": 0.1835271120071411,
"step": 1300
},
{
"epoch": 0.4192,
"grad_norm": 0.4893503189086914,
"learning_rate": 3.6055129718806836e-06,
"loss": 0.17832870483398439,
"step": 1310
},
{
"epoch": 0.4224,
"grad_norm": 0.46311140060424805,
"learning_rate": 3.5803934860568134e-06,
"loss": 0.17421386241912842,
"step": 1320
},
{
"epoch": 0.4256,
"grad_norm": 0.5104085803031921,
"learning_rate": 3.5551391515986163e-06,
"loss": 0.17839887142181396,
"step": 1330
},
{
"epoch": 0.4288,
"grad_norm": 0.5179736018180847,
"learning_rate": 3.529753120609982e-06,
"loss": 0.17463579177856445,
"step": 1340
},
{
"epoch": 0.432,
"grad_norm": 0.5104801654815674,
"learning_rate": 3.5042385616324243e-06,
"loss": 0.17735633850097657,
"step": 1350
},
{
"epoch": 0.4352,
"grad_norm": 0.4137459993362427,
"learning_rate": 3.4785986592495934e-06,
"loss": 0.17638884782791137,
"step": 1360
},
{
"epoch": 0.4384,
"grad_norm": 0.40862590074539185,
"learning_rate": 3.452836613689803e-06,
"loss": 0.17429978847503663,
"step": 1370
},
{
"epoch": 0.4416,
"grad_norm": 0.4509046673774719,
"learning_rate": 3.426955640426584e-06,
"loss": 0.17711507081985473,
"step": 1380
},
{
"epoch": 0.4448,
"grad_norm": 0.41654732823371887,
"learning_rate": 3.4009589697773605e-06,
"loss": 0.17662335634231568,
"step": 1390
},
{
"epoch": 0.448,
"grad_norm": 0.44660893082618713,
"learning_rate": 3.3748498465002475e-06,
"loss": 0.17387230396270753,
"step": 1400
},
{
"epoch": 0.4512,
"grad_norm": 0.4556673765182495,
"learning_rate": 3.3486315293890693e-06,
"loss": 0.1813045024871826,
"step": 1410
},
{
"epoch": 0.4544,
"grad_norm": 0.47021520137786865,
"learning_rate": 3.3223072908666053e-06,
"loss": 0.17260375022888183,
"step": 1420
},
{
"epoch": 0.4576,
"grad_norm": 0.47382044792175293,
"learning_rate": 3.295880416576153e-06,
"loss": 0.1822422504425049,
"step": 1430
},
{
"epoch": 0.4608,
"grad_norm": 0.4591139554977417,
"learning_rate": 3.269354204971427e-06,
"loss": 0.17569501399993898,
"step": 1440
},
{
"epoch": 0.464,
"grad_norm": 0.5069236755371094,
"learning_rate": 3.242731966904865e-06,
"loss": 0.1741647243499756,
"step": 1450
},
{
"epoch": 0.4672,
"grad_norm": 0.47500520944595337,
"learning_rate": 3.2160170252143913e-06,
"loss": 0.17435843944549562,
"step": 1460
},
{
"epoch": 0.4704,
"grad_norm": 0.4182279109954834,
"learning_rate": 3.1892127143086716e-06,
"loss": 0.18048869371414183,
"step": 1470
},
{
"epoch": 0.4736,
"grad_norm": 0.4673804044723511,
"learning_rate": 3.1623223797509347e-06,
"loss": 0.1697669506072998,
"step": 1480
},
{
"epoch": 0.4768,
"grad_norm": 0.5240358710289001,
"learning_rate": 3.135349377841396e-06,
"loss": 0.17352311611175536,
"step": 1490
},
{
"epoch": 0.48,
"grad_norm": 0.4815457761287689,
"learning_rate": 3.1082970751983497e-06,
"loss": 0.17982523441314696,
"step": 1500
},
{
"epoch": 0.4832,
"grad_norm": 0.4510243237018585,
"learning_rate": 3.0811688483379546e-06,
"loss": 0.17709922790527344,
"step": 1510
},
{
"epoch": 0.4864,
"grad_norm": 0.49320361018180847,
"learning_rate": 3.0539680832528074e-06,
"loss": 0.17739441394805908,
"step": 1520
},
{
"epoch": 0.4896,
"grad_norm": 0.500535786151886,
"learning_rate": 3.026698174989316e-06,
"loss": 0.17496352195739745,
"step": 1530
},
{
"epoch": 0.4928,
"grad_norm": 0.4047980010509491,
"learning_rate": 2.999362527223952e-06,
"loss": 0.17673254013061523,
"step": 1540
},
{
"epoch": 0.496,
"grad_norm": 0.44109296798706055,
"learning_rate": 2.9719645518384194e-06,
"loss": 0.17737388610839844,
"step": 1550
},
{
"epoch": 0.4992,
"grad_norm": 0.5294535160064697,
"learning_rate": 2.944507668493807e-06,
"loss": 0.17971434593200683,
"step": 1560
},
{
"epoch": 0.5024,
"grad_norm": 0.4643715023994446,
"learning_rate": 2.9169953042037623e-06,
"loss": 0.1756264567375183,
"step": 1570
},
{
"epoch": 0.5056,
"grad_norm": 0.5058485865592957,
"learning_rate": 2.889430892906754e-06,
"loss": 0.17225983142852783,
"step": 1580
},
{
"epoch": 0.5088,
"grad_norm": 0.44740837812423706,
"learning_rate": 2.861817875037462e-06,
"loss": 0.16964515447616577,
"step": 1590
},
{
"epoch": 0.512,
"grad_norm": 0.4732149839401245,
"learning_rate": 2.8341596970973683e-06,
"loss": 0.1788057804107666,
"step": 1600
},
{
"epoch": 0.5152,
"grad_norm": 0.4844909906387329,
"learning_rate": 2.80645981122458e-06,
"loss": 0.17410922050476074,
"step": 1610
},
{
"epoch": 0.5184,
"grad_norm": 0.42798951268196106,
"learning_rate": 2.7787216747629508e-06,
"loss": 0.17997238636016846,
"step": 1620
},
{
"epoch": 0.5216,
"grad_norm": 0.40051475167274475,
"learning_rate": 2.7509487498305615e-06,
"loss": 0.1678016424179077,
"step": 1630
},
{
"epoch": 0.5248,
"grad_norm": 0.4792901873588562,
"learning_rate": 2.7231445028875924e-06,
"loss": 0.17371201515197754,
"step": 1640
},
{
"epoch": 0.528,
"grad_norm": 0.47259077429771423,
"learning_rate": 2.6953124043036604e-06,
"loss": 0.1733928442001343,
"step": 1650
},
{
"epoch": 0.5312,
"grad_norm": 0.47142958641052246,
"learning_rate": 2.667455927924667e-06,
"loss": 0.1721155285835266,
"step": 1660
},
{
"epoch": 0.5344,
"grad_norm": 0.45087724924087524,
"learning_rate": 2.6395785506392164e-06,
"loss": 0.16819543838500978,
"step": 1670
},
{
"epoch": 0.5376,
"grad_norm": 0.46989184617996216,
"learning_rate": 2.6116837519446407e-06,
"loss": 0.17220993041992189,
"step": 1680
},
{
"epoch": 0.5408,
"grad_norm": 0.46657466888427734,
"learning_rate": 2.5837750135127192e-06,
"loss": 0.1694636106491089,
"step": 1690
},
{
"epoch": 0.544,
"grad_norm": 0.4564855098724365,
"learning_rate": 2.555855818755108e-06,
"loss": 0.17499985694885253,
"step": 1700
},
{
"epoch": 0.5472,
"grad_norm": 0.46723976731300354,
"learning_rate": 2.5279296523885636e-06,
"loss": 0.17200368642807007,
"step": 1710
},
{
"epoch": 0.5504,
"grad_norm": 0.4736442565917969,
"learning_rate": 2.5e-06,
"loss": 0.1722201704978943,
"step": 1720
},
{
"epoch": 0.5536,
"grad_norm": 0.46455293893814087,
"learning_rate": 2.472070347611437e-06,
"loss": 0.16756889820098878,
"step": 1730
},
{
"epoch": 0.5568,
"grad_norm": 0.5148719549179077,
"learning_rate": 2.444144181244893e-06,
"loss": 0.1717504858970642,
"step": 1740
},
{
"epoch": 0.56,
"grad_norm": 0.4863338768482208,
"learning_rate": 2.416224986487282e-06,
"loss": 0.16999526023864747,
"step": 1750
},
{
"epoch": 0.5632,
"grad_norm": 0.4933525323867798,
"learning_rate": 2.3883162480553605e-06,
"loss": 0.167264187335968,
"step": 1760
},
{
"epoch": 0.5664,
"grad_norm": 0.47000402212142944,
"learning_rate": 2.3604214493607844e-06,
"loss": 0.17092803716659546,
"step": 1770
},
{
"epoch": 0.5696,
"grad_norm": 0.5010572075843811,
"learning_rate": 2.332544072075333e-06,
"loss": 0.17414600849151612,
"step": 1780
},
{
"epoch": 0.5728,
"grad_norm": 0.48513519763946533,
"learning_rate": 2.30468759569634e-06,
"loss": 0.17303409576416015,
"step": 1790
},
{
"epoch": 0.576,
"grad_norm": 0.5627573728561401,
"learning_rate": 2.276855497112408e-06,
"loss": 0.1702120542526245,
"step": 1800
},
{
"epoch": 0.5792,
"grad_norm": 0.4507666230201721,
"learning_rate": 2.2490512501694394e-06,
"loss": 0.16601570844650268,
"step": 1810
},
{
"epoch": 0.5824,
"grad_norm": 0.44019457697868347,
"learning_rate": 2.2212783252370496e-06,
"loss": 0.1702256679534912,
"step": 1820
},
{
"epoch": 0.5856,
"grad_norm": 0.4958936870098114,
"learning_rate": 2.1935401887754213e-06,
"loss": 0.17476747035980225,
"step": 1830
},
{
"epoch": 0.5888,
"grad_norm": 0.5116126537322998,
"learning_rate": 2.165840302902632e-06,
"loss": 0.16839592456817626,
"step": 1840
},
{
"epoch": 0.592,
"grad_norm": 0.45241907238960266,
"learning_rate": 2.1381821249625383e-06,
"loss": 0.17359941005706786,
"step": 1850
},
{
"epoch": 0.5952,
"grad_norm": 0.4567364454269409,
"learning_rate": 2.1105691070932465e-06,
"loss": 0.16584824323654174,
"step": 1860
},
{
"epoch": 0.5984,
"grad_norm": 0.4592301845550537,
"learning_rate": 2.083004695796238e-06,
"loss": 0.16465656757354735,
"step": 1870
},
{
"epoch": 0.6016,
"grad_norm": 0.5488863587379456,
"learning_rate": 2.055492331506194e-06,
"loss": 0.17121386528015137,
"step": 1880
},
{
"epoch": 0.6048,
"grad_norm": 0.5324817299842834,
"learning_rate": 2.0280354481615814e-06,
"loss": 0.16972928047180175,
"step": 1890
},
{
"epoch": 0.608,
"grad_norm": 0.48892301321029663,
"learning_rate": 2.000637472776049e-06,
"loss": 0.170430588722229,
"step": 1900
},
{
"epoch": 0.6112,
"grad_norm": 0.47859957814216614,
"learning_rate": 1.973301825010685e-06,
"loss": 0.16521625518798827,
"step": 1910
},
{
"epoch": 0.6144,
"grad_norm": 0.4869942367076874,
"learning_rate": 1.9460319167471934e-06,
"loss": 0.17017598152160646,
"step": 1920
},
{
"epoch": 0.6176,
"grad_norm": 0.45271286368370056,
"learning_rate": 1.9188311516620466e-06,
"loss": 0.16556847095489502,
"step": 1930
},
{
"epoch": 0.6208,
"grad_norm": 0.5558905005455017,
"learning_rate": 1.891702924801651e-06,
"loss": 0.1715184211730957,
"step": 1940
},
{
"epoch": 0.624,
"grad_norm": 0.585498571395874,
"learning_rate": 1.864650622158604e-06,
"loss": 0.17473173141479492,
"step": 1950
},
{
"epoch": 0.6272,
"grad_norm": 0.48184657096862793,
"learning_rate": 1.8376776202490666e-06,
"loss": 0.17099432945251464,
"step": 1960
},
{
"epoch": 0.6304,
"grad_norm": 0.5236788392066956,
"learning_rate": 1.8107872856913293e-06,
"loss": 0.16600199937820434,
"step": 1970
},
{
"epoch": 0.6336,
"grad_norm": 0.46146464347839355,
"learning_rate": 1.7839829747856096e-06,
"loss": 0.1678165912628174,
"step": 1980
},
{
"epoch": 0.6368,
"grad_norm": 0.5520542860031128,
"learning_rate": 1.7572680330951359e-06,
"loss": 0.1660862684249878,
"step": 1990
},
{
"epoch": 0.64,
"grad_norm": 0.4780367910861969,
"learning_rate": 1.7306457950285747e-06,
"loss": 0.16768529415130615,
"step": 2000
},
{
"epoch": 0.6432,
"grad_norm": 0.49715733528137207,
"learning_rate": 1.704119583423848e-06,
"loss": 0.1622864842414856,
"step": 2010
},
{
"epoch": 0.6464,
"grad_norm": 0.4588998258113861,
"learning_rate": 1.677692709133396e-06,
"loss": 0.16991891860961914,
"step": 2020
},
{
"epoch": 0.6496,
"grad_norm": 0.515424370765686,
"learning_rate": 1.6513684706109311e-06,
"loss": 0.1692219614982605,
"step": 2030
},
{
"epoch": 0.6528,
"grad_norm": 0.5444751977920532,
"learning_rate": 1.6251501534997529e-06,
"loss": 0.16774747371673585,
"step": 2040
},
{
"epoch": 0.656,
"grad_norm": 0.46245628595352173,
"learning_rate": 1.5990410302226405e-06,
"loss": 0.1620822787284851,
"step": 2050
},
{
"epoch": 0.6592,
"grad_norm": 0.4891432225704193,
"learning_rate": 1.5730443595734162e-06,
"loss": 0.16800422668457032,
"step": 2060
},
{
"epoch": 0.6624,
"grad_norm": 0.5363055467605591,
"learning_rate": 1.5471633863101982e-06,
"loss": 0.16863477230072021,
"step": 2070
},
{
"epoch": 0.6656,
"grad_norm": 0.5734866261482239,
"learning_rate": 1.521401340750407e-06,
"loss": 0.17034238576889038,
"step": 2080
},
{
"epoch": 0.6688,
"grad_norm": 0.5337269902229309,
"learning_rate": 1.495761438367577e-06,
"loss": 0.1681105375289917,
"step": 2090
},
{
"epoch": 0.672,
"grad_norm": 0.553776204586029,
"learning_rate": 1.4702468793900187e-06,
"loss": 0.16727843284606933,
"step": 2100
},
{
"epoch": 0.6752,
"grad_norm": 0.4766635298728943,
"learning_rate": 1.444860848401384e-06,
"loss": 0.16690752506256104,
"step": 2110
},
{
"epoch": 0.6784,
"grad_norm": 0.5310425758361816,
"learning_rate": 1.4196065139431866e-06,
"loss": 0.166944682598114,
"step": 2120
},
{
"epoch": 0.6816,
"grad_norm": 0.5216221809387207,
"learning_rate": 1.3944870281193178e-06,
"loss": 0.16600677967071534,
"step": 2130
},
{
"epoch": 0.6848,
"grad_norm": 0.48818936944007874,
"learning_rate": 1.3695055262026208e-06,
"loss": 0.17084674835205077,
"step": 2140
},
{
"epoch": 0.688,
"grad_norm": 0.4620889723300934,
"learning_rate": 1.3446651262435679e-06,
"loss": 0.16466886997222902,
"step": 2150
},
{
"epoch": 0.6912,
"grad_norm": 0.5135729908943176,
"learning_rate": 1.3199689286810746e-06,
"loss": 0.16638584136962892,
"step": 2160
},
{
"epoch": 0.6944,
"grad_norm": 0.5088804364204407,
"learning_rate": 1.2954200159555294e-06,
"loss": 0.16140960454940795,
"step": 2170
},
{
"epoch": 0.6976,
"grad_norm": 0.5873313546180725,
"learning_rate": 1.2710214521240527e-06,
"loss": 0.16632808446884156,
"step": 2180
},
{
"epoch": 0.7008,
"grad_norm": 0.5511561036109924,
"learning_rate": 1.246776282478063e-06,
"loss": 0.16666927337646484,
"step": 2190
},
{
"epoch": 0.704,
"grad_norm": 0.5059501528739929,
"learning_rate": 1.222687533163181e-06,
"loss": 0.1660354971885681,
"step": 2200
},
{
"epoch": 0.7072,
"grad_norm": 0.5848804116249084,
"learning_rate": 1.1987582108015228e-06,
"loss": 0.167069149017334,
"step": 2210
},
{
"epoch": 0.7104,
"grad_norm": 0.551862359046936,
"learning_rate": 1.1749913021164255e-06,
"loss": 0.1673675775527954,
"step": 2220
},
{
"epoch": 0.7136,
"grad_norm": 0.545039713382721,
"learning_rate": 1.1513897735596702e-06,
"loss": 0.16290701627731324,
"step": 2230
},
{
"epoch": 0.7168,
"grad_norm": 0.5339580178260803,
"learning_rate": 1.127956570941218e-06,
"loss": 0.16006081104278563,
"step": 2240
},
{
"epoch": 0.72,
"grad_norm": 0.6216872930526733,
"learning_rate": 1.104694619061533e-06,
"loss": 0.16769601106643678,
"step": 2250
},
{
"epoch": 0.7232,
"grad_norm": 0.5120308995246887,
"learning_rate": 1.0816068213465295e-06,
"loss": 0.1700352430343628,
"step": 2260
},
{
"epoch": 0.7264,
"grad_norm": 0.5662424564361572,
"learning_rate": 1.0586960594851762e-06,
"loss": 0.16522903442382814,
"step": 2270
},
{
"epoch": 0.7296,
"grad_norm": 0.5819669365882874,
"learning_rate": 1.0359651930698217e-06,
"loss": 0.16397686004638673,
"step": 2280
},
{
"epoch": 0.7328,
"grad_norm": 0.5186004042625427,
"learning_rate": 1.0134170592392837e-06,
"loss": 0.1659336805343628,
"step": 2290
},
{
"epoch": 0.736,
"grad_norm": 0.548119843006134,
"learning_rate": 9.910544723247204e-07,
"loss": 0.16470273733139038,
"step": 2300
},
{
"epoch": 0.7392,
"grad_norm": 0.5775285959243774,
"learning_rate": 9.688802234983706e-07,
"loss": 0.16365277767181396,
"step": 2310
},
{
"epoch": 0.7424,
"grad_norm": 0.527587354183197,
"learning_rate": 9.468970804251742e-07,
"loss": 0.1663814902305603,
"step": 2320
},
{
"epoch": 0.7456,
"grad_norm": 0.5380632281303406,
"learning_rate": 9.251077869173244e-07,
"loss": 0.1611353039741516,
"step": 2330
},
{
"epoch": 0.7488,
"grad_norm": 0.5524181723594666,
"learning_rate": 9.035150625918054e-07,
"loss": 0.16146550178527833,
"step": 2340
},
{
"epoch": 0.752,
"grad_norm": 0.5604588985443115,
"learning_rate": 8.821216025309395e-07,
"loss": 0.16433074474334716,
"step": 2350
},
{
"epoch": 0.7552,
"grad_norm": 0.5228386521339417,
"learning_rate": 8.609300769460055e-07,
"loss": 0.16035985946655273,
"step": 2360
},
{
"epoch": 0.7584,
"grad_norm": 0.5645732283592224,
"learning_rate": 8.399431308439592e-07,
"loss": 0.16581802368164061,
"step": 2370
},
{
"epoch": 0.7616,
"grad_norm": 0.5424208045005798,
"learning_rate": 8.191633836972962e-07,
"loss": 0.16631256341934203,
"step": 2380
},
{
"epoch": 0.7648,
"grad_norm": 0.5168543457984924,
"learning_rate": 7.985934291171024e-07,
"loss": 0.16223517656326295,
"step": 2390
},
{
"epoch": 0.768,
"grad_norm": 0.5431832075119019,
"learning_rate": 7.7823583452934e-07,
"loss": 0.16580779552459718,
"step": 2400
},
{
"epoch": 0.7712,
"grad_norm": 0.5234625339508057,
"learning_rate": 7.58093140854389e-07,
"loss": 0.1643224239349365,
"step": 2410
},
{
"epoch": 0.7744,
"grad_norm": 0.5630083680152893,
"learning_rate": 7.381678621899077e-07,
"loss": 0.16557644605636596,
"step": 2420
},
{
"epoch": 0.7776,
"grad_norm": 0.564468264579773,
"learning_rate": 7.184624854970379e-07,
"loss": 0.16128422021865846,
"step": 2430
},
{
"epoch": 0.7808,
"grad_norm": 0.5211976170539856,
"learning_rate": 6.989794702899932e-07,
"loss": 0.16088463068008424,
"step": 2440
},
{
"epoch": 0.784,
"grad_norm": 0.5524035096168518,
"learning_rate": 6.797212483290777e-07,
"loss": 0.162975013256073,
"step": 2450
},
{
"epoch": 0.7872,
"grad_norm": 0.5299164652824402,
"learning_rate": 6.60690223317171e-07,
"loss": 0.1605682849884033,
"step": 2460
},
{
"epoch": 0.7904,
"grad_norm": 0.5172830820083618,
"learning_rate": 6.418887705997046e-07,
"loss": 0.16314079761505126,
"step": 2470
},
{
"epoch": 0.7936,
"grad_norm": 0.600565493106842,
"learning_rate": 6.23319236868189e-07,
"loss": 0.16519222259521485,
"step": 2480
},
{
"epoch": 0.7968,
"grad_norm": 0.5615212321281433,
"learning_rate": 6.049839398673141e-07,
"loss": 0.1658394455909729,
"step": 2490
},
{
"epoch": 0.8,
"grad_norm": 0.6055536866188049,
"learning_rate": 5.868851681056567e-07,
"loss": 0.1668398857116699,
"step": 2500
},
{
"epoch": 0.8032,
"grad_norm": 0.5374464988708496,
"learning_rate": 5.690251805700467e-07,
"loss": 0.1650017738342285,
"step": 2510
},
{
"epoch": 0.8064,
"grad_norm": 0.5940120816230774,
"learning_rate": 5.514062064436096e-07,
"loss": 0.16656217575073243,
"step": 2520
},
{
"epoch": 0.8096,
"grad_norm": 0.5781455039978027,
"learning_rate": 5.34030444827533e-07,
"loss": 0.16284786462783812,
"step": 2530
},
{
"epoch": 0.8128,
"grad_norm": 0.6245594024658203,
"learning_rate": 5.169000644665895e-07,
"loss": 0.16080342531204223,
"step": 2540
},
{
"epoch": 0.816,
"grad_norm": 0.5611685514450073,
"learning_rate": 5.000172034784442e-07,
"loss": 0.16152651309967042,
"step": 2550
},
{
"epoch": 0.8192,
"grad_norm": 0.6154001951217651,
"learning_rate": 4.833839690867853e-07,
"loss": 0.16308718919754028,
"step": 2560
},
{
"epoch": 0.8224,
"grad_norm": 0.603412926197052,
"learning_rate": 4.6700243735831705e-07,
"loss": 0.15839451551437378,
"step": 2570
},
{
"epoch": 0.8256,
"grad_norm": 0.5827690362930298,
"learning_rate": 4.508746529436311e-07,
"loss": 0.15733330249786376,
"step": 2580
},
{
"epoch": 0.8288,
"grad_norm": 0.5977171659469604,
"learning_rate": 4.350026288220083e-07,
"loss": 0.1626259684562683,
"step": 2590
},
{
"epoch": 0.832,
"grad_norm": 0.577282726764679,
"learning_rate": 4.1938834605017133e-07,
"loss": 0.16123565435409545,
"step": 2600
},
{
"epoch": 0.8352,
"grad_norm": 0.5709326267242432,
"learning_rate": 4.0403375351501515e-07,
"loss": 0.1605008363723755,
"step": 2610
},
{
"epoch": 0.8384,
"grad_norm": 0.5334956049919128,
"learning_rate": 3.88940767690362e-07,
"loss": 0.15991599559783937,
"step": 2620
},
{
"epoch": 0.8416,
"grad_norm": 0.6113538146018982,
"learning_rate": 3.7411127239775774e-07,
"loss": 0.15835078954696655,
"step": 2630
},
{
"epoch": 0.8448,
"grad_norm": 0.5511420965194702,
"learning_rate": 3.595471185713431e-07,
"loss": 0.15640071630477906,
"step": 2640
},
{
"epoch": 0.848,
"grad_norm": 0.6026161313056946,
"learning_rate": 3.4525012402682826e-07,
"loss": 0.1582810401916504,
"step": 2650
},
{
"epoch": 0.8512,
"grad_norm": 0.584636926651001,
"learning_rate": 3.3122207323460804e-07,
"loss": 0.16068333387374878,
"step": 2660
},
{
"epoch": 0.8544,
"grad_norm": 0.6109904646873474,
"learning_rate": 3.1746471709702963e-07,
"loss": 0.15707263946533204,
"step": 2670
},
{
"epoch": 0.8576,
"grad_norm": 0.5890913009643555,
"learning_rate": 3.039797727298585e-07,
"loss": 0.16340034008026122,
"step": 2680
},
{
"epoch": 0.8608,
"grad_norm": 0.5972010493278503,
"learning_rate": 2.9076892324795546e-07,
"loss": 0.16199719905853271,
"step": 2690
},
{
"epoch": 0.864,
"grad_norm": 0.5684933662414551,
"learning_rate": 2.778338175551995e-07,
"loss": 0.157791805267334,
"step": 2700
},
{
"epoch": 0.8672,
"grad_norm": 0.5997552871704102,
"learning_rate": 2.6517607013868326e-07,
"loss": 0.16547198295593263,
"step": 2710
},
{
"epoch": 0.8704,
"grad_norm": 0.5953125953674316,
"learning_rate": 2.527972608672002e-07,
"loss": 0.16187076568603515,
"step": 2720
},
{
"epoch": 0.8736,
"grad_norm": 0.6352548003196716,
"learning_rate": 2.40698934794053e-07,
"loss": 0.16053141355514527,
"step": 2730
},
{
"epoch": 0.8768,
"grad_norm": 0.5708346366882324,
"learning_rate": 2.2888260196421237e-07,
"loss": 0.15960177183151245,
"step": 2740
},
{
"epoch": 0.88,
"grad_norm": 0.5586206912994385,
"learning_rate": 2.1734973722583735e-07,
"loss": 0.16245728731155396,
"step": 2750
},
{
"epoch": 0.8832,
"grad_norm": 0.568911075592041,
"learning_rate": 2.0610178004619564e-07,
"loss": 0.15456781387329102,
"step": 2760
},
{
"epoch": 0.8864,
"grad_norm": 0.5839951634407043,
"learning_rate": 1.9514013433199834e-07,
"loss": 0.1657383680343628,
"step": 2770
},
{
"epoch": 0.8896,
"grad_norm": 0.5618033409118652,
"learning_rate": 1.8446616825416958e-07,
"loss": 0.16385490894317628,
"step": 2780
},
{
"epoch": 0.8928,
"grad_norm": 0.6637345552444458,
"learning_rate": 1.7408121407708007e-07,
"loss": 0.1651144504547119,
"step": 2790
},
{
"epoch": 0.896,
"grad_norm": 0.6217195391654968,
"learning_rate": 1.6398656799226253e-07,
"loss": 0.15437800884246827,
"step": 2800
},
{
"epoch": 0.8992,
"grad_norm": 0.5852035284042358,
"learning_rate": 1.5418348995662773e-07,
"loss": 0.15989196300506592,
"step": 2810
},
{
"epoch": 0.9024,
"grad_norm": 0.5942526459693909,
"learning_rate": 1.4467320353520275e-07,
"loss": 0.15796263217926027,
"step": 2820
},
{
"epoch": 0.9056,
"grad_norm": 0.6098470687866211,
"learning_rate": 1.3545689574841341e-07,
"loss": 0.15871069431304932,
"step": 2830
},
{
"epoch": 0.9088,
"grad_norm": 0.5966947674751282,
"learning_rate": 1.26535716923927e-07,
"loss": 0.16476703882217408,
"step": 2840
},
{
"epoch": 0.912,
"grad_norm": 0.6142219305038452,
"learning_rate": 1.1791078055307493e-07,
"loss": 0.16100226640701293,
"step": 2850
},
{
"epoch": 0.9152,
"grad_norm": 0.6414586901664734,
"learning_rate": 1.0958316315187289e-07,
"loss": 0.1599841833114624,
"step": 2860
},
{
"epoch": 0.9184,
"grad_norm": 0.5591761469841003,
"learning_rate": 1.0155390412665528e-07,
"loss": 0.15885504484176635,
"step": 2870
},
{
"epoch": 0.9216,
"grad_norm": 0.5783108472824097,
"learning_rate": 9.38240056443443e-08,
"loss": 0.1566369891166687,
"step": 2880
},
{
"epoch": 0.9248,
"grad_norm": 0.5923556685447693,
"learning_rate": 8.639443250736402e-08,
"loss": 0.15874269008636474,
"step": 2890
},
{
"epoch": 0.928,
"grad_norm": 0.5997049808502197,
"learning_rate": 7.926611203321777e-08,
"loss": 0.16008044481277467,
"step": 2900
},
{
"epoch": 0.9312,
"grad_norm": 0.6039618253707886,
"learning_rate": 7.243993393874882e-08,
"loss": 0.16028298139572145,
"step": 2910
},
{
"epoch": 0.9344,
"grad_norm": 0.5951699614524841,
"learning_rate": 6.591675022908805e-08,
"loss": 0.1579365015029907,
"step": 2920
},
{
"epoch": 0.9376,
"grad_norm": 0.5689706206321716,
"learning_rate": 5.969737509131241e-08,
"loss": 0.15579373836517335,
"step": 2930
},
{
"epoch": 0.9408,
"grad_norm": 0.6160743236541748,
"learning_rate": 5.3782584792823334e-08,
"loss": 0.16048473119735718,
"step": 2940
},
{
"epoch": 0.944,
"grad_norm": 0.6045230627059937,
"learning_rate": 4.817311758445686e-08,
"loss": 0.16096792221069336,
"step": 2950
},
{
"epoch": 0.9472,
"grad_norm": 0.5729262232780457,
"learning_rate": 4.286967360833866e-08,
"loss": 0.16088333129882812,
"step": 2960
},
{
"epoch": 0.9504,
"grad_norm": 0.5658770203590393,
"learning_rate": 3.787291481049754e-08,
"loss": 0.16108319759368897,
"step": 2970
},
{
"epoch": 0.9536,
"grad_norm": 0.6044265627861023,
"learning_rate": 3.3183464858244364e-08,
"loss": 0.16624315977096557,
"step": 2980
},
{
"epoch": 0.9568,
"grad_norm": 0.5959712862968445,
"learning_rate": 2.8801909062328992e-08,
"loss": 0.15484490394592285,
"step": 2990
},
{
"epoch": 0.96,
"grad_norm": 0.61944979429245,
"learning_rate": 2.4728794303886248e-08,
"loss": 0.15291309356689453,
"step": 3000
},
{
"epoch": 0.9632,
"grad_norm": 0.6033022403717041,
"learning_rate": 2.0964628966175794e-08,
"loss": 0.15846710205078124,
"step": 3010
},
{
"epoch": 0.9664,
"grad_norm": 0.5712878704071045,
"learning_rate": 1.750988287113009e-08,
"loss": 0.15711405277252197,
"step": 3020
},
{
"epoch": 0.9696,
"grad_norm": 0.5986247062683105,
"learning_rate": 1.4364987220713278e-08,
"loss": 0.16440168619155884,
"step": 3030
},
{
"epoch": 0.9728,
"grad_norm": 0.6022250652313232,
"learning_rate": 1.1530334543099763e-08,
"loss": 0.16256611347198485,
"step": 3040
},
{
"epoch": 0.976,
"grad_norm": 0.6291612386703491,
"learning_rate": 9.006278643683697e-09,
"loss": 0.1630788564682007,
"step": 3050
},
{
"epoch": 0.9792,
"grad_norm": 0.6009607315063477,
"learning_rate": 6.793134560916514e-09,
"loss": 0.16497414112091063,
"step": 3060
},
{
"epoch": 0.9824,
"grad_norm": 0.5608439445495605,
"learning_rate": 4.891178526986451e-09,
"loss": 0.1616500735282898,
"step": 3070
},
{
"epoch": 0.9856,
"grad_norm": 0.5863735675811768,
"learning_rate": 3.3006479333413943e-09,
"loss": 0.16401275396347045,
"step": 3080
},
{
"epoch": 0.9888,
"grad_norm": 0.5719097256660461,
"learning_rate": 2.021741301058422e-09,
"loss": 0.16051123142242432,
"step": 3090
},
{
"epoch": 0.992,
"grad_norm": 0.5933088064193726,
"learning_rate": 1.0546182560652872e-09,
"loss": 0.16212167739868164,
"step": 3100
},
{
"epoch": 0.9952,
"grad_norm": 0.5979989171028137,
"learning_rate": 3.9939950921774607e-10,
"loss": 0.1578517436981201,
"step": 3110
},
{
"epoch": 0.9984,
"grad_norm": 0.5368886590003967,
"learning_rate": 5.616684123160854e-11,
"loss": 0.15727875232696534,
"step": 3120
},
{
"epoch": 1.0,
"step": 3125,
"total_flos": 2.0482865290660545e+19,
"train_loss": 0.18024104360580445,
"train_runtime": 72884.8323,
"train_samples_per_second": 2.744,
"train_steps_per_second": 0.043
}
],
"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": 2.0482865290660545e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}