| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1497, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02004008016032064, | |
| "grad_norm": 11.386945789449237, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.914, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04008016032064128, | |
| "grad_norm": 3.686249022148282, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.7761, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06012024048096192, | |
| "grad_norm": 1.797608399405574, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.6765, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08016032064128256, | |
| "grad_norm": 2.53598544592724, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.6319, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10020040080160321, | |
| "grad_norm": 1.8597121806159485, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.6028, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12024048096192384, | |
| "grad_norm": 1.8378706472462616, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.5846, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1402805611222445, | |
| "grad_norm": 4.6666512736387356, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.5761, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 2.67064642378821, | |
| "learning_rate": 4.999862725880242e-06, | |
| "loss": 0.5758, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18036072144288579, | |
| "grad_norm": 2.6004134353948736, | |
| "learning_rate": 4.998764633422446e-06, | |
| "loss": 0.5667, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20040080160320642, | |
| "grad_norm": 2.0205968573263413, | |
| "learning_rate": 4.996568984453794e-06, | |
| "loss": 0.5651, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22044088176352705, | |
| "grad_norm": 1.7372715757375135, | |
| "learning_rate": 4.993276850606589e-06, | |
| "loss": 0.5593, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24048096192384769, | |
| "grad_norm": 2.3534996172184846, | |
| "learning_rate": 4.988889838675462e-06, | |
| "loss": 0.5566, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2605210420841683, | |
| "grad_norm": 2.1712051144800846, | |
| "learning_rate": 4.9834100898331425e-06, | |
| "loss": 0.554, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.280561122244489, | |
| "grad_norm": 2.988740945183358, | |
| "learning_rate": 4.976840278585413e-06, | |
| "loss": 0.5412, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30060120240480964, | |
| "grad_norm": 1.7513590918597028, | |
| "learning_rate": 4.969183611465766e-06, | |
| "loss": 0.5494, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 1.911642443149266, | |
| "learning_rate": 4.9604438254703845e-06, | |
| "loss": 0.5342, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3406813627254509, | |
| "grad_norm": 1.7028624872037597, | |
| "learning_rate": 4.95062518623422e-06, | |
| "loss": 0.5427, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36072144288577157, | |
| "grad_norm": 1.89229572623421, | |
| "learning_rate": 4.939732485949066e-06, | |
| "loss": 0.5379, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3807615230460922, | |
| "grad_norm": 1.458083413953667, | |
| "learning_rate": 4.927771041024627e-06, | |
| "loss": 0.5355, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.40080160320641284, | |
| "grad_norm": 2.1624686975408336, | |
| "learning_rate": 4.914746689493731e-06, | |
| "loss": 0.54, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42084168336673344, | |
| "grad_norm": 1.841839502355017, | |
| "learning_rate": 4.900665788162962e-06, | |
| "loss": 0.5306, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4408817635270541, | |
| "grad_norm": 1.5623403972638494, | |
| "learning_rate": 4.885535209510081e-06, | |
| "loss": 0.531, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.46092184368737477, | |
| "grad_norm": 2.0959167807238672, | |
| "learning_rate": 4.869362338329773e-06, | |
| "loss": 0.528, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 2.877452797335848, | |
| "learning_rate": 4.852155068129343e-06, | |
| "loss": 0.53, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.501002004008016, | |
| "grad_norm": 2.406716714855783, | |
| "learning_rate": 4.8339217972761274e-06, | |
| "loss": 0.5242, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5210420841683366, | |
| "grad_norm": 2.6180161170191956, | |
| "learning_rate": 4.814671424898493e-06, | |
| "loss": 0.5218, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5410821643286573, | |
| "grad_norm": 1.9325754069366325, | |
| "learning_rate": 4.794413346542432e-06, | |
| "loss": 0.516, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.561122244488978, | |
| "grad_norm": 1.6202018849080733, | |
| "learning_rate": 4.773157449585871e-06, | |
| "loss": 0.519, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5811623246492986, | |
| "grad_norm": 2.5386712058137846, | |
| "learning_rate": 4.750914108412927e-06, | |
| "loss": 0.5168, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6012024048096193, | |
| "grad_norm": 1.6204343506402092, | |
| "learning_rate": 4.727694179350476e-06, | |
| "loss": 0.5125, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6212424849699398, | |
| "grad_norm": 1.6404070884775066, | |
| "learning_rate": 4.703508995369497e-06, | |
| "loss": 0.513, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 1.6188579894187152, | |
| "learning_rate": 4.6783703605537715e-06, | |
| "loss": 0.5147, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6613226452905812, | |
| "grad_norm": 1.4143375173363102, | |
| "learning_rate": 4.652290544338663e-06, | |
| "loss": 0.5173, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6813627254509018, | |
| "grad_norm": 2.468479009235932, | |
| "learning_rate": 4.6252822755227605e-06, | |
| "loss": 0.5108, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7014028056112225, | |
| "grad_norm": 2.1271900619659854, | |
| "learning_rate": 4.5973587360553185e-06, | |
| "loss": 0.5131, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7214428857715431, | |
| "grad_norm": 1.5534652608564319, | |
| "learning_rate": 4.568533554602535e-06, | |
| "loss": 0.5083, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7414829659318637, | |
| "grad_norm": 1.6027013883759695, | |
| "learning_rate": 4.5388207998957986e-06, | |
| "loss": 0.5133, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7615230460921844, | |
| "grad_norm": 1.530068712838613, | |
| "learning_rate": 4.508234973865145e-06, | |
| "loss": 0.5157, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.781563126252505, | |
| "grad_norm": 1.8195363489136889, | |
| "learning_rate": 4.476791004561293e-06, | |
| "loss": 0.512, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 1.637599921361452, | |
| "learning_rate": 4.4445042388696935e-06, | |
| "loss": 0.5095, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8216432865731463, | |
| "grad_norm": 1.5903329614208361, | |
| "learning_rate": 4.411390435020164e-06, | |
| "loss": 0.5131, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8416833667334669, | |
| "grad_norm": 1.5488153255590384, | |
| "learning_rate": 4.377465754895757e-06, | |
| "loss": 0.5052, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8617234468937875, | |
| "grad_norm": 1.8543008230812608, | |
| "learning_rate": 4.342746756144609e-06, | |
| "loss": 0.5124, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8817635270541082, | |
| "grad_norm": 1.5858389816681295, | |
| "learning_rate": 4.307250384098645e-06, | |
| "loss": 0.5094, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9018036072144289, | |
| "grad_norm": 1.6693267301390629, | |
| "learning_rate": 4.270993963503048e-06, | |
| "loss": 0.5055, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9218436873747495, | |
| "grad_norm": 1.4123291299839993, | |
| "learning_rate": 4.233995190060559e-06, | |
| "loss": 0.511, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9418837675350702, | |
| "grad_norm": 1.4294865023874812, | |
| "learning_rate": 4.196272121794714e-06, | |
| "loss": 0.5122, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 1.4747480031112472, | |
| "learning_rate": 4.15784317023626e-06, | |
| "loss": 0.5068, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9819639278557114, | |
| "grad_norm": 1.6382019572026507, | |
| "learning_rate": 4.118727091437013e-06, | |
| "loss": 0.5002, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5069113373756409, | |
| "eval_runtime": 45.884, | |
| "eval_samples_per_second": 292.695, | |
| "eval_steps_per_second": 1.155, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.002004008016032, | |
| "grad_norm": 3.2862336764098803, | |
| "learning_rate": 4.078942976815581e-06, | |
| "loss": 0.4942, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0220440881763526, | |
| "grad_norm": 1.7759252107171999, | |
| "learning_rate": 4.038510243839396e-06, | |
| "loss": 0.418, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0420841683366733, | |
| "grad_norm": 1.7448507912514557, | |
| "learning_rate": 3.997448626547613e-06, | |
| "loss": 0.4136, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.062124248496994, | |
| "grad_norm": 1.4081466868123829, | |
| "learning_rate": 3.955778165919502e-06, | |
| "loss": 0.4131, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0821643286573146, | |
| "grad_norm": 1.4737485149847567, | |
| "learning_rate": 3.913519200093035e-06, | |
| "loss": 0.412, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1022044088176353, | |
| "grad_norm": 1.2635878378102765, | |
| "learning_rate": 3.870692354438423e-06, | |
| "loss": 0.4142, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.122244488977956, | |
| "grad_norm": 1.553384214976449, | |
| "learning_rate": 3.827318531491478e-06, | |
| "loss": 0.4115, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1422845691382766, | |
| "grad_norm": 1.5870294004406946, | |
| "learning_rate": 3.7834189007517002e-06, | |
| "loss": 0.4108, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.1623246492985972, | |
| "grad_norm": 1.6029534467749516, | |
| "learning_rate": 3.739014888350053e-06, | |
| "loss": 0.4116, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.182364729458918, | |
| "grad_norm": 1.6194120236052227, | |
| "learning_rate": 3.694128166591494e-06, | |
| "loss": 0.4136, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2024048096192386, | |
| "grad_norm": 1.436406692065181, | |
| "learning_rate": 3.6487806433773615e-06, | |
| "loss": 0.4134, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2224448897795592, | |
| "grad_norm": 1.4163845777049167, | |
| "learning_rate": 3.6029944515127585e-06, | |
| "loss": 0.4091, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2424849699398797, | |
| "grad_norm": 1.5714505381282495, | |
| "learning_rate": 3.5567919379041783e-06, | |
| "loss": 0.4142, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2625250501002003, | |
| "grad_norm": 1.4088322992815865, | |
| "learning_rate": 3.510195652652629e-06, | |
| "loss": 0.4161, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.282565130260521, | |
| "grad_norm": 1.6997351047753615, | |
| "learning_rate": 3.463228338047589e-06, | |
| "loss": 0.4059, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.3026052104208417, | |
| "grad_norm": 1.5683685074334606, | |
| "learning_rate": 3.4159129174671534e-06, | |
| "loss": 0.4065, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3226452905811623, | |
| "grad_norm": 1.6253512472033467, | |
| "learning_rate": 3.3682724841898067e-06, | |
| "loss": 0.4108, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.342685370741483, | |
| "grad_norm": 1.759831168192353, | |
| "learning_rate": 3.320330290123261e-06, | |
| "loss": 0.4103, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3627254509018036, | |
| "grad_norm": 1.7378970495344055, | |
| "learning_rate": 3.2721097344558794e-06, | |
| "loss": 0.4133, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3827655310621243, | |
| "grad_norm": 1.5427360104428394, | |
| "learning_rate": 3.223634352236213e-06, | |
| "loss": 0.4103, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.402805611222445, | |
| "grad_norm": 1.4316640042632263, | |
| "learning_rate": 3.1749278028862325e-06, | |
| "loss": 0.4078, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4228456913827654, | |
| "grad_norm": 1.951831574889627, | |
| "learning_rate": 3.1260138586538483e-06, | |
| "loss": 0.4108, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4428857715430863, | |
| "grad_norm": 1.5703450558182246, | |
| "learning_rate": 3.076916393010373e-06, | |
| "loss": 0.4091, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4629258517034067, | |
| "grad_norm": 1.4374885381526987, | |
| "learning_rate": 3.0276593689985737e-06, | |
| "loss": 0.4056, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.4829659318637274, | |
| "grad_norm": 1.2643689737278019, | |
| "learning_rate": 2.9782668275370095e-06, | |
| "loss": 0.4139, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.503006012024048, | |
| "grad_norm": 1.570266686325652, | |
| "learning_rate": 2.928762875686358e-06, | |
| "loss": 0.4109, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.5230460921843687, | |
| "grad_norm": 1.5743968516846465, | |
| "learning_rate": 2.879171674883462e-06, | |
| "loss": 0.4103, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5430861723446894, | |
| "grad_norm": 1.5844459981668437, | |
| "learning_rate": 2.8295174291488383e-06, | |
| "loss": 0.4109, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.56312625250501, | |
| "grad_norm": 1.504354777149479, | |
| "learning_rate": 2.779824373273397e-06, | |
| "loss": 0.4118, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.5831663326653307, | |
| "grad_norm": 1.3787147420821566, | |
| "learning_rate": 2.7301167609901474e-06, | |
| "loss": 0.4135, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6032064128256514, | |
| "grad_norm": 1.4777711687133863, | |
| "learning_rate": 2.680418853136659e-06, | |
| "loss": 0.4095, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.623246492985972, | |
| "grad_norm": 1.468500779993122, | |
| "learning_rate": 2.630754905814048e-06, | |
| "loss": 0.4079, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6432865731462925, | |
| "grad_norm": 1.6134930461454124, | |
| "learning_rate": 2.58114915854829e-06, | |
| "loss": 0.4083, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.6633266533066133, | |
| "grad_norm": 1.4588892173060275, | |
| "learning_rate": 2.531625822459603e-06, | |
| "loss": 0.4087, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.6833667334669338, | |
| "grad_norm": 1.4497521826737287, | |
| "learning_rate": 2.4822090684457147e-06, | |
| "loss": 0.4113, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.7034068136272547, | |
| "grad_norm": 1.4372338693007887, | |
| "learning_rate": 2.432923015384743e-06, | |
| "loss": 0.4047, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.723446893787575, | |
| "grad_norm": 1.3561893796274846, | |
| "learning_rate": 2.3837917183634815e-06, | |
| "loss": 0.4085, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.7434869739478958, | |
| "grad_norm": 1.3339197747345177, | |
| "learning_rate": 2.3348391569368064e-06, | |
| "loss": 0.411, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7635270541082164, | |
| "grad_norm": 1.2878261372500692, | |
| "learning_rate": 2.2860892234239565e-06, | |
| "loss": 0.4145, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.783567134268537, | |
| "grad_norm": 1.3852796037018127, | |
| "learning_rate": 2.237565711247391e-06, | |
| "loss": 0.4109, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.8036072144288577, | |
| "grad_norm": 1.343576505451422, | |
| "learning_rate": 2.1892923033199075e-06, | |
| "loss": 0.4078, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.8236472945891784, | |
| "grad_norm": 1.4910844256604878, | |
| "learning_rate": 2.141292560485708e-06, | |
| "loss": 0.4064, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.843687374749499, | |
| "grad_norm": 1.452632980241925, | |
| "learning_rate": 2.0935899100210316e-06, | |
| "loss": 0.4059, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.8637274549098195, | |
| "grad_norm": 1.3416125112736146, | |
| "learning_rate": 2.046207634199989e-06, | |
| "loss": 0.405, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.8837675350701404, | |
| "grad_norm": 1.3346098098736352, | |
| "learning_rate": 1.9991688589311575e-06, | |
| "loss": 0.4067, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.9038076152304608, | |
| "grad_norm": 1.2974237604973982, | |
| "learning_rate": 1.9524965424705026e-06, | |
| "loss": 0.4024, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.9238476953907817, | |
| "grad_norm": 1.471473412639078, | |
| "learning_rate": 1.9062134642161198e-06, | |
| "loss": 0.4091, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.9438877755511021, | |
| "grad_norm": 1.357694886526054, | |
| "learning_rate": 1.8603422135902737e-06, | |
| "loss": 0.4057, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.9639278557114228, | |
| "grad_norm": 1.2865807531711169, | |
| "learning_rate": 1.8149051790141628e-06, | |
| "loss": 0.4066, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.9839679358717435, | |
| "grad_norm": 1.297389613332899, | |
| "learning_rate": 1.7699245369807778e-06, | |
| "loss": 0.4045, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.4974905848503113, | |
| "eval_runtime": 51.1615, | |
| "eval_samples_per_second": 262.502, | |
| "eval_steps_per_second": 1.036, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 2.004008016032064, | |
| "grad_norm": 2.171277446089175, | |
| "learning_rate": 1.725422241231205e-06, | |
| "loss": 0.3859, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.024048096192385, | |
| "grad_norm": 1.7817202202370321, | |
| "learning_rate": 1.6814200120396438e-06, | |
| "loss": 0.3244, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.0440881763527052, | |
| "grad_norm": 1.619953212777582, | |
| "learning_rate": 1.6379393256123737e-06, | |
| "loss": 0.3211, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.064128256513026, | |
| "grad_norm": 1.40219382175259, | |
| "learning_rate": 1.595001403605844e-06, | |
| "loss": 0.3158, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.0841683366733466, | |
| "grad_norm": 1.4533085140998225, | |
| "learning_rate": 1.552627202769e-06, | |
| "loss": 0.318, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.1042084168336674, | |
| "grad_norm": 1.4269652142649647, | |
| "learning_rate": 1.5108374047149061e-06, | |
| "loss": 0.3197, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.124248496993988, | |
| "grad_norm": 1.537954200218747, | |
| "learning_rate": 1.4696524058266516e-06, | |
| "loss": 0.3196, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.1442885771543088, | |
| "grad_norm": 1.6449193889553193, | |
| "learning_rate": 1.4290923073024712e-06, | |
| "loss": 0.3214, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.164328657314629, | |
| "grad_norm": 1.3598251114998061, | |
| "learning_rate": 1.3891769053449355e-06, | |
| "loss": 0.321, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.18436873747495, | |
| "grad_norm": 1.551617792162964, | |
| "learning_rate": 1.3499256814990003e-06, | |
| "loss": 0.3172, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.2044088176352705, | |
| "grad_norm": 1.487413687096442, | |
| "learning_rate": 1.3113577931436332e-06, | |
| "loss": 0.319, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.2244488977955914, | |
| "grad_norm": 1.379093893132986, | |
| "learning_rate": 1.2734920641416556e-06, | |
| "loss": 0.3201, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.244488977955912, | |
| "grad_norm": 1.3525463916397538, | |
| "learning_rate": 1.236346975652358e-06, | |
| "loss": 0.32, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.2645290581162323, | |
| "grad_norm": 1.4572131883779704, | |
| "learning_rate": 1.1999406571113962e-06, | |
| "loss": 0.3171, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.284569138276553, | |
| "grad_norm": 1.4042689852309367, | |
| "learning_rate": 1.1642908773823274e-06, | |
| "loss": 0.3166, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.3046092184368736, | |
| "grad_norm": 1.3000414229131876, | |
| "learning_rate": 1.1294150360841577e-06, | |
| "loss": 0.3211, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.3246492985971945, | |
| "grad_norm": 1.3811842558284932, | |
| "learning_rate": 1.0953301550990882e-06, | |
| "loss": 0.3213, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.344689378757515, | |
| "grad_norm": 1.3259540554288098, | |
| "learning_rate": 1.0620528702646312e-06, | |
| "loss": 0.3193, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.364729458917836, | |
| "grad_norm": 1.4720294318649223, | |
| "learning_rate": 1.0295994232541454e-06, | |
| "loss": 0.32, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.3847695390781563, | |
| "grad_norm": 1.3734699044137106, | |
| "learning_rate": 9.979856536497435e-07, | |
| "loss": 0.3178, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.404809619238477, | |
| "grad_norm": 1.3553774328605976, | |
| "learning_rate": 9.672269912114582e-07, | |
| "loss": 0.3134, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.4248496993987976, | |
| "grad_norm": 1.350392284981299, | |
| "learning_rate": 9.373384483464223e-07, | |
| "loss": 0.3218, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.4448897795591185, | |
| "grad_norm": 1.4402104791397123, | |
| "learning_rate": 9.08334612781753e-07, | |
| "loss": 0.3205, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.464929859719439, | |
| "grad_norm": 1.3646686545789422, | |
| "learning_rate": 8.80229640444705e-07, | |
| "loss": 0.3216, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.4849699398797593, | |
| "grad_norm": 1.3499126545275342, | |
| "learning_rate": 8.530372485535751e-07, | |
| "loss": 0.3197, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.50501002004008, | |
| "grad_norm": 1.4013520657650143, | |
| "learning_rate": 8.267707089227288e-07, | |
| "loss": 0.3174, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.5250501002004007, | |
| "grad_norm": 1.3543143703928378, | |
| "learning_rate": 8.01442841485013e-07, | |
| "loss": 0.3237, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.5450901803607215, | |
| "grad_norm": 1.3528190597320051, | |
| "learning_rate": 7.770660080347213e-07, | |
| "loss": 0.3226, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.565130260521042, | |
| "grad_norm": 1.316933143733213, | |
| "learning_rate": 7.536521061941613e-07, | |
| "loss": 0.3178, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.585170340681363, | |
| "grad_norm": 1.2849910299224803, | |
| "learning_rate": 7.312125636067723e-07, | |
| "loss": 0.3225, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.6052104208416833, | |
| "grad_norm": 1.290228394422378, | |
| "learning_rate": 7.097583323596257e-07, | |
| "loss": 0.3222, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.6252505010020037, | |
| "grad_norm": 1.426441845965089, | |
| "learning_rate": 6.892998836380285e-07, | |
| "loss": 0.3173, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.6452905811623246, | |
| "grad_norm": 1.2801979621928115, | |
| "learning_rate": 6.698472026148459e-07, | |
| "loss": 0.3234, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.6653306613226455, | |
| "grad_norm": 1.3220972865520855, | |
| "learning_rate": 6.514097835770269e-07, | |
| "loss": 0.3175, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.685370741482966, | |
| "grad_norm": 1.3667409352018853, | |
| "learning_rate": 6.3399662529172e-07, | |
| "loss": 0.3184, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.7054108216432864, | |
| "grad_norm": 1.3528891456726284, | |
| "learning_rate": 6.176162266142376e-07, | |
| "loss": 0.323, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.7254509018036073, | |
| "grad_norm": 1.327321152396037, | |
| "learning_rate": 6.022765823400116e-07, | |
| "loss": 0.3215, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.7454909819639277, | |
| "grad_norm": 1.277409240062218, | |
| "learning_rate": 5.879851793025669e-07, | |
| "loss": 0.3182, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.7655310621242486, | |
| "grad_norm": 1.336386040490389, | |
| "learning_rate": 5.747489927194165e-07, | |
| "loss": 0.3202, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.785571142284569, | |
| "grad_norm": 1.2533448631397004, | |
| "learning_rate": 5.625744827876601e-07, | |
| "loss": 0.319, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.80561122244489, | |
| "grad_norm": 1.290234810317301, | |
| "learning_rate": 5.514675915309507e-07, | |
| "loss": 0.3185, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.8256513026052104, | |
| "grad_norm": 1.311881104807796, | |
| "learning_rate": 5.414337398993661e-07, | |
| "loss": 0.3156, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.845691382765531, | |
| "grad_norm": 1.3245343587549412, | |
| "learning_rate": 5.324778251236008e-07, | |
| "loss": 0.3191, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.8657314629258517, | |
| "grad_norm": 1.318165777680666, | |
| "learning_rate": 5.246042183247698e-07, | |
| "loss": 0.3141, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.8857715430861726, | |
| "grad_norm": 1.3833599862766017, | |
| "learning_rate": 5.178167623809932e-07, | |
| "loss": 0.3199, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.905811623246493, | |
| "grad_norm": 1.3572877385241544, | |
| "learning_rate": 5.121187700517977e-07, | |
| "loss": 0.3213, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.9258517034068134, | |
| "grad_norm": 1.289697605126265, | |
| "learning_rate": 5.075130223612563e-07, | |
| "loss": 0.3209, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.9458917835671343, | |
| "grad_norm": 1.3713215738345175, | |
| "learning_rate": 5.040017672406508e-07, | |
| "loss": 0.3152, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.9659318637274548, | |
| "grad_norm": 1.3065474715872867, | |
| "learning_rate": 5.015867184313224e-07, | |
| "loss": 0.3152, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.9859719438877756, | |
| "grad_norm": 1.29789945819745, | |
| "learning_rate": 5.002690546482441e-07, | |
| "loss": 0.3184, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.5188336968421936, | |
| "eval_runtime": 52.0293, | |
| "eval_samples_per_second": 258.124, | |
| "eval_steps_per_second": 1.019, | |
| "step": 1497 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1497, | |
| "total_flos": 2507530756423680.0, | |
| "train_loss": 0.4253670667598625, | |
| "train_runtime": 9374.4269, | |
| "train_samples_per_second": 81.656, | |
| "train_steps_per_second": 0.16 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1497, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 2507530756423680.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |