| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.7961476725521669, | |
| "eval_steps": 150, | |
| "global_step": 155, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005136436597110754, | |
| "grad_norm": 104.2032241821289, | |
| "learning_rate": 2e-05, | |
| "loss": 43.8361, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005136436597110754, | |
| "eval_loss": 2.7993953227996826, | |
| "eval_runtime": 137.9974, | |
| "eval_samples_per_second": 4.754, | |
| "eval_steps_per_second": 1.188, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010272873194221509, | |
| "grad_norm": 87.10350799560547, | |
| "learning_rate": 4e-05, | |
| "loss": 45.8657, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.015409309791332263, | |
| "grad_norm": 96.3443374633789, | |
| "learning_rate": 6e-05, | |
| "loss": 43.7377, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.020545746388443017, | |
| "grad_norm": 94.10945892333984, | |
| "learning_rate": 8e-05, | |
| "loss": 40.9281, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.025682182985553772, | |
| "grad_norm": 92.76588439941406, | |
| "learning_rate": 0.0001, | |
| "loss": 33.8946, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.030818619582664526, | |
| "grad_norm": 53.15927505493164, | |
| "learning_rate": 0.00012, | |
| "loss": 24.9538, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.035955056179775284, | |
| "grad_norm": 84.37348937988281, | |
| "learning_rate": 0.00014, | |
| "loss": 23.256, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.041091492776886035, | |
| "grad_norm": 32.134864807128906, | |
| "learning_rate": 0.00016, | |
| "loss": 21.5102, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04622792937399679, | |
| "grad_norm": 28.912691116333008, | |
| "learning_rate": 0.00018, | |
| "loss": 17.9462, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.051364365971107544, | |
| "grad_norm": 23.97195053100586, | |
| "learning_rate": 0.0002, | |
| "loss": 16.3874, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0565008025682183, | |
| "grad_norm": 53.18326187133789, | |
| "learning_rate": 0.00019997652980184843, | |
| "loss": 18.0917, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06163723916532905, | |
| "grad_norm": 31.358110427856445, | |
| "learning_rate": 0.0001999061302243977, | |
| "loss": 17.1273, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06677367576243981, | |
| "grad_norm": 27.875253677368164, | |
| "learning_rate": 0.00019978883431348845, | |
| "loss": 15.7283, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07191011235955057, | |
| "grad_norm": 29.975252151489258, | |
| "learning_rate": 0.00019962469712828614, | |
| "loss": 16.687, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.07704654895666131, | |
| "grad_norm": 35.4009895324707, | |
| "learning_rate": 0.00019941379571543596, | |
| "loss": 17.2859, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08218298555377207, | |
| "grad_norm": 89.57384490966797, | |
| "learning_rate": 0.00019915622907289694, | |
| "loss": 17.1492, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.08731942215088283, | |
| "grad_norm": 35.824459075927734, | |
| "learning_rate": 0.00019885211810347184, | |
| "loss": 14.816, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.09245585874799359, | |
| "grad_norm": 21.76598358154297, | |
| "learning_rate": 0.00019850160555805486, | |
| "loss": 14.582, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.09759229534510433, | |
| "grad_norm": 22.013967514038086, | |
| "learning_rate": 0.00019810485596862392, | |
| "loss": 13.9424, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.10272873194221509, | |
| "grad_norm": 19.30277442932129, | |
| "learning_rate": 0.00019766205557100868, | |
| "loss": 15.0135, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.10786516853932585, | |
| "grad_norm": 20.07377052307129, | |
| "learning_rate": 0.00019717341221747056, | |
| "loss": 15.026, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1130016051364366, | |
| "grad_norm": 19.19417953491211, | |
| "learning_rate": 0.00019663915527913625, | |
| "loss": 16.28, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.11813804173354735, | |
| "grad_norm": 145.2245330810547, | |
| "learning_rate": 0.00019605953553832988, | |
| "loss": 14.1316, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1232744783306581, | |
| "grad_norm": 21.75243377685547, | |
| "learning_rate": 0.00019543482507085482, | |
| "loss": 15.4915, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.12841091492776885, | |
| "grad_norm": 19.737768173217773, | |
| "learning_rate": 0.00019476531711828027, | |
| "loss": 15.0702, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.13354735152487962, | |
| "grad_norm": 21.004024505615234, | |
| "learning_rate": 0.0001940513259502924, | |
| "loss": 13.8939, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.13868378812199036, | |
| "grad_norm": 99.0382080078125, | |
| "learning_rate": 0.0001932931867171751, | |
| "loss": 14.1338, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.14382022471910114, | |
| "grad_norm": 21.052730560302734, | |
| "learning_rate": 0.0001924912552924889, | |
| "loss": 15.3611, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.14895666131621188, | |
| "grad_norm": 25.83721160888672, | |
| "learning_rate": 0.00019164590810602262, | |
| "loss": 14.6511, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.15409309791332262, | |
| "grad_norm": 23.324485778808594, | |
| "learning_rate": 0.00019075754196709572, | |
| "loss": 13.7324, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1592295345104334, | |
| "grad_norm": 17.926301956176758, | |
| "learning_rate": 0.00018982657387829445, | |
| "loss": 14.6006, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.16436597110754414, | |
| "grad_norm": 22.068103790283203, | |
| "learning_rate": 0.00018885344083972914, | |
| "loss": 15.3865, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.16950240770465488, | |
| "grad_norm": 20.398998260498047, | |
| "learning_rate": 0.00018783859964390464, | |
| "loss": 14.2294, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.17463884430176566, | |
| "grad_norm": 17.491390228271484, | |
| "learning_rate": 0.00018678252666130013, | |
| "loss": 14.1286, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1797752808988764, | |
| "grad_norm": 19.122509002685547, | |
| "learning_rate": 0.00018568571761675893, | |
| "loss": 15.3119, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.18491171749598717, | |
| "grad_norm": 18.384124755859375, | |
| "learning_rate": 0.0001845486873567932, | |
| "loss": 15.4398, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.19004815409309792, | |
| "grad_norm": 41.315956115722656, | |
| "learning_rate": 0.00018337196960791302, | |
| "loss": 14.5058, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.19518459069020866, | |
| "grad_norm": 19.567920684814453, | |
| "learning_rate": 0.00018215611672609317, | |
| "loss": 12.7428, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.20032102728731943, | |
| "grad_norm": 19.43627166748047, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 15.1849, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.20545746388443017, | |
| "grad_norm": 49.411319732666016, | |
| "learning_rate": 0.00017960930657056438, | |
| "loss": 13.3985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.21059390048154092, | |
| "grad_norm": 21.388917922973633, | |
| "learning_rate": 0.00017827954477963557, | |
| "loss": 14.1779, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2157303370786517, | |
| "grad_norm": 24.790122985839844, | |
| "learning_rate": 0.0001769130382601629, | |
| "loss": 15.1373, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.22086677367576243, | |
| "grad_norm": 16.753856658935547, | |
| "learning_rate": 0.00017551042845572208, | |
| "loss": 14.7309, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2260032102728732, | |
| "grad_norm": 23.899765014648438, | |
| "learning_rate": 0.00017407237375691392, | |
| "loss": 13.6093, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.23113964686998395, | |
| "grad_norm": 52.38229751586914, | |
| "learning_rate": 0.0001725995491923131, | |
| "loss": 14.5637, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2362760834670947, | |
| "grad_norm": 44.96168518066406, | |
| "learning_rate": 0.00017109264611160708, | |
| "loss": 14.0165, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.24141252006420547, | |
| "grad_norm": 20.429908752441406, | |
| "learning_rate": 0.00016955237186107387, | |
| "loss": 14.7569, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2465489566613162, | |
| "grad_norm": 16.509567260742188, | |
| "learning_rate": 0.0001679794494515508, | |
| "loss": 13.9921, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.251685393258427, | |
| "grad_norm": 16.397178649902344, | |
| "learning_rate": 0.00016637461721905045, | |
| "loss": 13.2361, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2568218298555377, | |
| "grad_norm": 16.84309196472168, | |
| "learning_rate": 0.00016473862847818277, | |
| "loss": 15.3742, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.26195826645264847, | |
| "grad_norm": 17.62554359436035, | |
| "learning_rate": 0.00016307225116854622, | |
| "loss": 14.4287, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.26709470304975924, | |
| "grad_norm": 17.205963134765625, | |
| "learning_rate": 0.00016137626749425377, | |
| "loss": 13.9601, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.27223113964686996, | |
| "grad_norm": 19.4569091796875, | |
| "learning_rate": 0.00015965147355676343, | |
| "loss": 16.3653, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.27736757624398073, | |
| "grad_norm": 17.375333786010742, | |
| "learning_rate": 0.0001578986789811849, | |
| "loss": 14.8869, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2825040128410915, | |
| "grad_norm": 19.079904556274414, | |
| "learning_rate": 0.00015611870653623825, | |
| "loss": 13.38, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2876404494382023, | |
| "grad_norm": 14.686327934265137, | |
| "learning_rate": 0.00015431239174804328, | |
| "loss": 13.8367, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.292776886035313, | |
| "grad_norm": 16.750314712524414, | |
| "learning_rate": 0.00015248058250792008, | |
| "loss": 13.7871, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.29791332263242376, | |
| "grad_norm": 20.95732879638672, | |
| "learning_rate": 0.0001506241386743854, | |
| "loss": 13.5049, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.30304975922953453, | |
| "grad_norm": 678.4100341796875, | |
| "learning_rate": 0.00014874393166953192, | |
| "loss": 11.7639, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.30818619582664525, | |
| "grad_norm": 18.715354919433594, | |
| "learning_rate": 0.00014684084406997903, | |
| "loss": 11.5865, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.313322632423756, | |
| "grad_norm": 27.259925842285156, | |
| "learning_rate": 0.00014491576919258792, | |
| "loss": 16.0604, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3184590690208668, | |
| "grad_norm": 146.95001220703125, | |
| "learning_rate": 0.0001429696106751352, | |
| "loss": 15.6741, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.3235955056179775, | |
| "grad_norm": 16.70216941833496, | |
| "learning_rate": 0.0001410032820521416, | |
| "loss": 15.1472, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3287319422150883, | |
| "grad_norm": 24.176897048950195, | |
| "learning_rate": 0.00013901770632605547, | |
| "loss": 15.0085, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.33386837881219905, | |
| "grad_norm": 26.137481689453125, | |
| "learning_rate": 0.00013701381553399145, | |
| "loss": 14.074, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.33900481540930977, | |
| "grad_norm": 16.90602684020996, | |
| "learning_rate": 0.00013499255031022885, | |
| "loss": 13.6073, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.34414125200642054, | |
| "grad_norm": 15.70071029663086, | |
| "learning_rate": 0.00013295485944467405, | |
| "loss": 13.4469, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.3492776886035313, | |
| "grad_norm": 16.358449935913086, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 13.356, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.354414125200642, | |
| "grad_norm": 39.20793914794922, | |
| "learning_rate": 0.0001288340340501351, | |
| "loss": 13.7233, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3595505617977528, | |
| "grad_norm": 16.230451583862305, | |
| "learning_rate": 0.00012675283385292212, | |
| "loss": 12.6883, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.36468699839486357, | |
| "grad_norm": 17.057575225830078, | |
| "learning_rate": 0.00012465907576947622, | |
| "loss": 13.5811, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.36982343499197434, | |
| "grad_norm": 18.70912742614746, | |
| "learning_rate": 0.00012255374261813944, | |
| "loss": 17.5856, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.37495987158908506, | |
| "grad_norm": 17.773834228515625, | |
| "learning_rate": 0.0001204378226506365, | |
| "loss": 15.1836, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.38009630818619583, | |
| "grad_norm": 19.240266799926758, | |
| "learning_rate": 0.00011831230908818563, | |
| "loss": 14.201, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.3852327447833066, | |
| "grad_norm": 28.190780639648438, | |
| "learning_rate": 0.0001161781996552765, | |
| "loss": 14.8009, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3903691813804173, | |
| "grad_norm": 13.847818374633789, | |
| "learning_rate": 0.00011403649611133444, | |
| "loss": 11.7056, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3955056179775281, | |
| "grad_norm": 19.09795379638672, | |
| "learning_rate": 0.00011188820378049065, | |
| "loss": 14.7104, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.40064205457463886, | |
| "grad_norm": 16.784597396850586, | |
| "learning_rate": 0.00010973433107967902, | |
| "loss": 13.9271, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4057784911717496, | |
| "grad_norm": 18.32082748413086, | |
| "learning_rate": 0.00010757588904528106, | |
| "loss": 14.7108, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.41091492776886035, | |
| "grad_norm": 19.074981689453125, | |
| "learning_rate": 0.00010541389085854176, | |
| "loss": 16.207, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4160513643659711, | |
| "grad_norm": 18.795635223388672, | |
| "learning_rate": 0.00010324935136997806, | |
| "loss": 13.4865, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.42118780096308184, | |
| "grad_norm": 18.05921173095703, | |
| "learning_rate": 0.000101083286623004, | |
| "loss": 13.5321, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4263242375601926, | |
| "grad_norm": 26.4294490814209, | |
| "learning_rate": 9.891671337699602e-05, | |
| "loss": 14.5322, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4314606741573034, | |
| "grad_norm": 19.195995330810547, | |
| "learning_rate": 9.675064863002196e-05, | |
| "loss": 14.3342, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.43659711075441415, | |
| "grad_norm": 56.876407623291016, | |
| "learning_rate": 9.458610914145826e-05, | |
| "loss": 14.4958, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.44173354735152487, | |
| "grad_norm": 38.66791915893555, | |
| "learning_rate": 9.242411095471897e-05, | |
| "loss": 15.1417, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.44686998394863564, | |
| "grad_norm": 15.607385635375977, | |
| "learning_rate": 9.026566892032105e-05, | |
| "loss": 12.8892, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4520064205457464, | |
| "grad_norm": 18.23895263671875, | |
| "learning_rate": 8.811179621950936e-05, | |
| "loss": 11.6303, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 20.267276763916016, | |
| "learning_rate": 8.596350388866558e-05, | |
| "loss": 15.0553, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.4622792937399679, | |
| "grad_norm": 15.288996696472168, | |
| "learning_rate": 8.382180034472353e-05, | |
| "loss": 11.5903, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.46741573033707867, | |
| "grad_norm": 17.56521224975586, | |
| "learning_rate": 8.168769091181438e-05, | |
| "loss": 14.486, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4725521669341894, | |
| "grad_norm": 18.872352600097656, | |
| "learning_rate": 7.956217734936353e-05, | |
| "loss": 11.8002, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.47768860353130016, | |
| "grad_norm": 16.709739685058594, | |
| "learning_rate": 7.744625738186059e-05, | |
| "loss": 13.3616, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.48282504012841093, | |
| "grad_norm": 16.364673614501953, | |
| "learning_rate": 7.534092423052381e-05, | |
| "loss": 14.2758, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.48796147672552165, | |
| "grad_norm": 15.938226699829102, | |
| "learning_rate": 7.324716614707793e-05, | |
| "loss": 11.5942, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.4930979133226324, | |
| "grad_norm": 17.955299377441406, | |
| "learning_rate": 7.116596594986494e-05, | |
| "loss": 14.6283, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.4982343499197432, | |
| "grad_norm": 19.787822723388672, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 14.1204, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.503370786516854, | |
| "grad_norm": 24.217247009277344, | |
| "learning_rate": 6.704514055532597e-05, | |
| "loss": 12.5203, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5085072231139647, | |
| "grad_norm": 18.735475540161133, | |
| "learning_rate": 6.500744968977116e-05, | |
| "loss": 13.2639, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5136436597110754, | |
| "grad_norm": 16.328588485717773, | |
| "learning_rate": 6.298618446600856e-05, | |
| "loss": 13.0953, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5187800963081862, | |
| "grad_norm": 17.875574111938477, | |
| "learning_rate": 6.0982293673944544e-05, | |
| "loss": 12.6593, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5239165329052969, | |
| "grad_norm": 16.03655242919922, | |
| "learning_rate": 5.899671794785839e-05, | |
| "loss": 13.7208, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5290529695024077, | |
| "grad_norm": 15.7937593460083, | |
| "learning_rate": 5.703038932486484e-05, | |
| "loss": 12.1358, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5341894060995185, | |
| "grad_norm": 15.46942138671875, | |
| "learning_rate": 5.5084230807412126e-05, | |
| "loss": 13.246, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5393258426966292, | |
| "grad_norm": 21.15059471130371, | |
| "learning_rate": 5.3159155930021e-05, | |
| "loss": 17.4358, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5444622792937399, | |
| "grad_norm": 16.775680541992188, | |
| "learning_rate": 5.12560683304681e-05, | |
| "loss": 15.4665, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5495987158908507, | |
| "grad_norm": 15.216927528381348, | |
| "learning_rate": 4.9375861325614606e-05, | |
| "loss": 14.233, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5547351524879615, | |
| "grad_norm": 16.067556381225586, | |
| "learning_rate": 4.751941749207995e-05, | |
| "loss": 13.9475, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5598715890850723, | |
| "grad_norm": 15.383379936218262, | |
| "learning_rate": 4.5687608251956714e-05, | |
| "loss": 13.1619, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.565008025682183, | |
| "grad_norm": 13.594917297363281, | |
| "learning_rate": 4.388129346376178e-05, | |
| "loss": 10.6763, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5701444622792937, | |
| "grad_norm": 16.273210525512695, | |
| "learning_rate": 4.210132101881516e-05, | |
| "loss": 13.5974, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.5752808988764045, | |
| "grad_norm": 14.688011169433594, | |
| "learning_rate": 4.034852644323661e-05, | |
| "loss": 12.487, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.5804173354735153, | |
| "grad_norm": 15.278275489807129, | |
| "learning_rate": 3.862373250574626e-05, | |
| "loss": 11.589, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.585553772070626, | |
| "grad_norm": 16.51677703857422, | |
| "learning_rate": 3.6927748831453836e-05, | |
| "loss": 14.3822, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5906902086677368, | |
| "grad_norm": 15.376001358032227, | |
| "learning_rate": 3.5261371521817244e-05, | |
| "loss": 12.6597, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5958266452648475, | |
| "grad_norm": 16.46743392944336, | |
| "learning_rate": 3.3625382780949574e-05, | |
| "loss": 12.9537, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6009630818619582, | |
| "grad_norm": 16.67145347595215, | |
| "learning_rate": 3.202055054844921e-05, | |
| "loss": 15.0622, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6060995184590691, | |
| "grad_norm": 17.57502555847168, | |
| "learning_rate": 3.0447628138926156e-05, | |
| "loss": 14.1264, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6112359550561798, | |
| "grad_norm": 14.330567359924316, | |
| "learning_rate": 2.890735388839295e-05, | |
| "loss": 13.1008, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6163723916532905, | |
| "grad_norm": 16.226436614990234, | |
| "learning_rate": 2.7400450807686938e-05, | |
| "loss": 13.5438, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6215088282504013, | |
| "grad_norm": 18.105459213256836, | |
| "learning_rate": 2.59276262430861e-05, | |
| "loss": 13.7602, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.626645264847512, | |
| "grad_norm": 15.712982177734375, | |
| "learning_rate": 2.4489571544277945e-05, | |
| "loss": 14.5214, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6317817014446228, | |
| "grad_norm": 16.994945526123047, | |
| "learning_rate": 2.308696173983711e-05, | |
| "loss": 12.3241, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6369181380417336, | |
| "grad_norm": 23.144821166992188, | |
| "learning_rate": 2.1720455220364444e-05, | |
| "loss": 12.649, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6420545746388443, | |
| "grad_norm": 20.007465362548828, | |
| "learning_rate": 2.0390693429435627e-05, | |
| "loss": 12.3556, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.647191011235955, | |
| "grad_norm": 17.01409149169922, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 14.2751, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.6523274478330658, | |
| "grad_norm": 20.14617347717285, | |
| "learning_rate": 1.784388327390687e-05, | |
| "loss": 11.4152, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6574638844301766, | |
| "grad_norm": 15.188875198364258, | |
| "learning_rate": 1.6628030392087e-05, | |
| "loss": 11.5006, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6626003210272873, | |
| "grad_norm": 12.970837593078613, | |
| "learning_rate": 1.5451312643206827e-05, | |
| "loss": 10.4096, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.6677367576243981, | |
| "grad_norm": 23.07573890686035, | |
| "learning_rate": 1.4314282383241096e-05, | |
| "loss": 12.7064, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6728731942215088, | |
| "grad_norm": 15.414446830749512, | |
| "learning_rate": 1.3217473338699859e-05, | |
| "loss": 12.8737, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.6780096308186195, | |
| "grad_norm": 14.607954978942871, | |
| "learning_rate": 1.2161400356095375e-05, | |
| "loss": 12.8935, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.6831460674157304, | |
| "grad_norm": 16.611759185791016, | |
| "learning_rate": 1.1146559160270875e-05, | |
| "loss": 14.0584, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.6882825040128411, | |
| "grad_norm": 22.148141860961914, | |
| "learning_rate": 1.0173426121705576e-05, | |
| "loss": 12.2587, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.6934189406099518, | |
| "grad_norm": 18.16852378845215, | |
| "learning_rate": 9.242458032904311e-06, | |
| "loss": 14.9363, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.6985553772070626, | |
| "grad_norm": 22.371219635009766, | |
| "learning_rate": 8.354091893977401e-06, | |
| "loss": 14.528, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7036918138041733, | |
| "grad_norm": 13.901525497436523, | |
| "learning_rate": 7.508744707511117e-06, | |
| "loss": 14.1155, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.708828250401284, | |
| "grad_norm": 13.545099258422852, | |
| "learning_rate": 6.70681328282492e-06, | |
| "loss": 12.0036, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7139646869983949, | |
| "grad_norm": 13.885724067687988, | |
| "learning_rate": 5.948674049707603e-06, | |
| "loss": 11.6004, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7191011235955056, | |
| "grad_norm": 18.979543685913086, | |
| "learning_rate": 5.2346828817197655e-06, | |
| "loss": 12.442, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7242375601926164, | |
| "grad_norm": 18.40704917907715, | |
| "learning_rate": 4.565174929145188e-06, | |
| "loss": 13.0276, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7293739967897271, | |
| "grad_norm": 14.125924110412598, | |
| "learning_rate": 3.940464461670135e-06, | |
| "loss": 12.777, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7345104333868379, | |
| "grad_norm": 14.582460403442383, | |
| "learning_rate": 3.360844720863765e-06, | |
| "loss": 12.5393, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.7396468699839487, | |
| "grad_norm": 101.13643646240234, | |
| "learning_rate": 2.826587782529444e-06, | |
| "loss": 10.174, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7447833065810594, | |
| "grad_norm": 17.687610626220703, | |
| "learning_rate": 2.3379444289913342e-06, | |
| "loss": 14.5668, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7499197431781701, | |
| "grad_norm": 13.70815372467041, | |
| "learning_rate": 1.8951440313760837e-06, | |
| "loss": 12.749, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.755056179775281, | |
| "grad_norm": 15.379615783691406, | |
| "learning_rate": 1.4983944419451613e-06, | |
| "loss": 12.6604, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.7601926163723917, | |
| "grad_norm": 16.77601432800293, | |
| "learning_rate": 1.1478818965281911e-06, | |
| "loss": 15.1071, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.7653290529695024, | |
| "grad_norm": 17.265384674072266, | |
| "learning_rate": 8.437709271030603e-07, | |
| "loss": 13.4781, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.7704654895666132, | |
| "grad_norm": 16.5050106048584, | |
| "learning_rate": 5.862042845640403e-07, | |
| "loss": 12.4935, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7704654895666132, | |
| "eval_loss": 0.8121631145477295, | |
| "eval_runtime": 138.4812, | |
| "eval_samples_per_second": 4.737, | |
| "eval_steps_per_second": 1.184, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7756019261637239, | |
| "grad_norm": 15.610320091247559, | |
| "learning_rate": 3.7530287171387843e-07, | |
| "loss": 11.1469, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.7807383627608346, | |
| "grad_norm": 16.400800704956055, | |
| "learning_rate": 2.1116568651156076e-07, | |
| "loss": 14.1984, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.7858747993579455, | |
| "grad_norm": 14.4852933883667, | |
| "learning_rate": 9.386977560232879e-08, | |
| "loss": 12.8882, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.7910112359550562, | |
| "grad_norm": 17.637113571166992, | |
| "learning_rate": 2.347019815158724e-08, | |
| "loss": 15.6212, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.7961476725521669, | |
| "grad_norm": 16.612701416015625, | |
| "learning_rate": 0.0, | |
| "loss": 13.2727, | |
| "step": 155 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 155, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 2, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4215448129804698e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |