| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8635578583765112, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008635578583765112, | |
| "grad_norm": 0.6623008251190186, | |
| "learning_rate": 0.0002, | |
| "loss": 1.7783, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.017271157167530225, | |
| "grad_norm": 0.5188003182411194, | |
| "learning_rate": 0.00019994965423831854, | |
| "loss": 1.3036, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.025906735751295335, | |
| "grad_norm": 0.5021234154701233, | |
| "learning_rate": 0.00019979866764718843, | |
| "loss": 0.9562, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03454231433506045, | |
| "grad_norm": 0.4302387833595276, | |
| "learning_rate": 0.00019954719225730847, | |
| "loss": 0.9219, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04317789291882556, | |
| "grad_norm": 0.5690401792526245, | |
| "learning_rate": 0.00019919548128307954, | |
| "loss": 0.9422, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05181347150259067, | |
| "grad_norm": 0.5356280207633972, | |
| "learning_rate": 0.00019874388886763944, | |
| "loss": 0.8915, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06044905008635579, | |
| "grad_norm": 0.4261893332004547, | |
| "learning_rate": 0.00019819286972627066, | |
| "loss": 0.8435, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0690846286701209, | |
| "grad_norm": 0.5050117373466492, | |
| "learning_rate": 0.00019754297868854073, | |
| "loss": 0.8846, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07772020725388601, | |
| "grad_norm": 0.39581066370010376, | |
| "learning_rate": 0.00019679487013963564, | |
| "loss": 0.8297, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08635578583765112, | |
| "grad_norm": 0.45784541964530945, | |
| "learning_rate": 0.00019594929736144976, | |
| "loss": 0.8334, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09499136442141623, | |
| "grad_norm": 0.43146270513534546, | |
| "learning_rate": 0.00019500711177409454, | |
| "loss": 0.8614, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10362694300518134, | |
| "grad_norm": 0.38489606976509094, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.7923, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11226252158894647, | |
| "grad_norm": 0.4087812006473541, | |
| "learning_rate": 0.00019283679330160726, | |
| "loss": 0.8122, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.12089810017271158, | |
| "grad_norm": 0.579022228717804, | |
| "learning_rate": 0.00019161084574320696, | |
| "loss": 0.7754, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12953367875647667, | |
| "grad_norm": 0.4284251928329468, | |
| "learning_rate": 0.00019029265382866214, | |
| "loss": 0.8205, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1381692573402418, | |
| "grad_norm": 0.4063330590724945, | |
| "learning_rate": 0.00018888354486549237, | |
| "loss": 0.8301, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14680483592400692, | |
| "grad_norm": 0.42471617460250854, | |
| "learning_rate": 0.00018738493770697852, | |
| "loss": 0.8253, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15544041450777202, | |
| "grad_norm": 0.44914358854293823, | |
| "learning_rate": 0.00018579834132349772, | |
| "loss": 0.8211, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.16407599309153714, | |
| "grad_norm": 0.35791847109794617, | |
| "learning_rate": 0.00018412535328311814, | |
| "loss": 0.8241, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.17271157167530224, | |
| "grad_norm": 0.41551169753074646, | |
| "learning_rate": 0.0001823676581429833, | |
| "loss": 0.8187, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18134715025906736, | |
| "grad_norm": 0.3767964243888855, | |
| "learning_rate": 0.00018052702575310588, | |
| "loss": 0.7815, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18998272884283246, | |
| "grad_norm": 0.3870415687561035, | |
| "learning_rate": 0.00017860530947427875, | |
| "loss": 0.7907, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.19861830742659758, | |
| "grad_norm": 0.4934289753437042, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.7788, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.20725388601036268, | |
| "grad_norm": 0.40702807903289795, | |
| "learning_rate": 0.0001745264449675755, | |
| "loss": 0.7932, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2158894645941278, | |
| "grad_norm": 0.40032950043678284, | |
| "learning_rate": 0.00017237340381050703, | |
| "loss": 0.7682, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.22452504317789293, | |
| "grad_norm": 0.4420917332172394, | |
| "learning_rate": 0.00017014748877063214, | |
| "loss": 0.7663, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.23316062176165803, | |
| "grad_norm": 0.3428107500076294, | |
| "learning_rate": 0.00016785094115571322, | |
| "loss": 0.7693, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.24179620034542315, | |
| "grad_norm": 0.7044374942779541, | |
| "learning_rate": 0.00016548607339452853, | |
| "loss": 0.7809, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2504317789291883, | |
| "grad_norm": 0.37226754426956177, | |
| "learning_rate": 0.00016305526670845226, | |
| "loss": 0.8263, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.25906735751295334, | |
| "grad_norm": 0.40080901980400085, | |
| "learning_rate": 0.00016056096871376667, | |
| "loss": 0.8182, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.26770293609671847, | |
| "grad_norm": 0.37441566586494446, | |
| "learning_rate": 0.00015800569095711982, | |
| "loss": 0.7597, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2763385146804836, | |
| "grad_norm": 0.4305630624294281, | |
| "learning_rate": 0.00015539200638661104, | |
| "loss": 0.801, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2849740932642487, | |
| "grad_norm": 0.482138991355896, | |
| "learning_rate": 0.00015272254676105025, | |
| "loss": 0.7541, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.29360967184801384, | |
| "grad_norm": 0.40154337882995605, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.8057, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3022452504317789, | |
| "grad_norm": 0.3598850667476654, | |
| "learning_rate": 0.0001472271074772683, | |
| "loss": 0.756, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.31088082901554404, | |
| "grad_norm": 0.46537917852401733, | |
| "learning_rate": 0.00014440666126057744, | |
| "loss": 0.7563, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.31951640759930916, | |
| "grad_norm": 0.44803386926651, | |
| "learning_rate": 0.00014154150130018866, | |
| "loss": 0.7908, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3281519861830743, | |
| "grad_norm": 0.4054509401321411, | |
| "learning_rate": 0.00013863451256931287, | |
| "loss": 0.7768, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.33678756476683935, | |
| "grad_norm": 0.3694852292537689, | |
| "learning_rate": 0.00013568862215918717, | |
| "loss": 0.7508, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3454231433506045, | |
| "grad_norm": 0.44693148136138916, | |
| "learning_rate": 0.00013270679633174218, | |
| "loss": 0.7723, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3540587219343696, | |
| "grad_norm": 0.47046658396720886, | |
| "learning_rate": 0.0001296920375328275, | |
| "loss": 0.7662, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3626943005181347, | |
| "grad_norm": 0.4129233658313751, | |
| "learning_rate": 0.00012664738136900348, | |
| "loss": 0.7546, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.37132987910189985, | |
| "grad_norm": 0.34191545844078064, | |
| "learning_rate": 0.00012357589355094275, | |
| "loss": 0.7586, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3799654576856649, | |
| "grad_norm": 0.37853655219078064, | |
| "learning_rate": 0.00012048066680651908, | |
| "loss": 0.7964, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.38860103626943004, | |
| "grad_norm": 0.4268665313720703, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.758, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.39723661485319517, | |
| "grad_norm": 0.3980403542518616, | |
| "learning_rate": 0.00011423148382732853, | |
| "loss": 0.7618, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4058721934369603, | |
| "grad_norm": 0.3414579927921295, | |
| "learning_rate": 0.00011108381999010111, | |
| "loss": 0.7867, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.41450777202072536, | |
| "grad_norm": 0.3817692697048187, | |
| "learning_rate": 0.00010792499568567884, | |
| "loss": 0.738, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4231433506044905, | |
| "grad_norm": 0.43348562717437744, | |
| "learning_rate": 0.00010475819158237425, | |
| "loss": 0.8304, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4317789291882556, | |
| "grad_norm": 0.31839361786842346, | |
| "learning_rate": 0.00010158659638348081, | |
| "loss": 0.7905, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.44041450777202074, | |
| "grad_norm": 0.3816024661064148, | |
| "learning_rate": 9.84134036165192e-05, | |
| "loss": 0.7286, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.44905008635578586, | |
| "grad_norm": 0.39195144176483154, | |
| "learning_rate": 9.524180841762577e-05, | |
| "loss": 0.7773, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.45768566493955093, | |
| "grad_norm": 0.4253169298171997, | |
| "learning_rate": 9.207500431432115e-05, | |
| "loss": 0.7905, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.46632124352331605, | |
| "grad_norm": 0.4840407967567444, | |
| "learning_rate": 8.891618000989891e-05, | |
| "loss": 0.7924, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4749568221070812, | |
| "grad_norm": 0.36509501934051514, | |
| "learning_rate": 8.57685161726715e-05, | |
| "loss": 0.772, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4835924006908463, | |
| "grad_norm": 0.3916187584400177, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.7499, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.49222797927461137, | |
| "grad_norm": 0.39139890670776367, | |
| "learning_rate": 7.951933319348095e-05, | |
| "loss": 0.7694, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5008635578583766, | |
| "grad_norm": 0.4077214300632477, | |
| "learning_rate": 7.642410644905726e-05, | |
| "loss": 0.7335, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5094991364421416, | |
| "grad_norm": 0.39956483244895935, | |
| "learning_rate": 7.335261863099651e-05, | |
| "loss": 0.7627, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5181347150259067, | |
| "grad_norm": 0.5192585587501526, | |
| "learning_rate": 7.030796246717255e-05, | |
| "loss": 0.7224, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5267702936096719, | |
| "grad_norm": 0.3494277000427246, | |
| "learning_rate": 6.729320366825784e-05, | |
| "loss": 0.7576, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5354058721934369, | |
| "grad_norm": 0.3863460123538971, | |
| "learning_rate": 6.431137784081282e-05, | |
| "loss": 0.7628, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5440414507772021, | |
| "grad_norm": 0.3836175501346588, | |
| "learning_rate": 6.136548743068713e-05, | |
| "loss": 0.7852, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5526770293609672, | |
| "grad_norm": 0.3461022973060608, | |
| "learning_rate": 5.845849869981137e-05, | |
| "loss": 0.7745, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5613126079447323, | |
| "grad_norm": 0.4282170534133911, | |
| "learning_rate": 5.559333873942259e-05, | |
| "loss": 0.7755, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5699481865284974, | |
| "grad_norm": 0.457685261964798, | |
| "learning_rate": 5.277289252273174e-05, | |
| "loss": 0.7358, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5785837651122625, | |
| "grad_norm": 0.37797123193740845, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.7377, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5872193436960277, | |
| "grad_norm": 0.3950762450695038, | |
| "learning_rate": 4.727745323894976e-05, | |
| "loss": 0.7795, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5958549222797928, | |
| "grad_norm": 0.3642215132713318, | |
| "learning_rate": 4.4607993613388976e-05, | |
| "loss": 0.7442, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6044905008635578, | |
| "grad_norm": 0.3860316574573517, | |
| "learning_rate": 4.19943090428802e-05, | |
| "loss": 0.7364, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.613126079447323, | |
| "grad_norm": 0.36184850335121155, | |
| "learning_rate": 3.943903128623335e-05, | |
| "loss": 0.7424, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6217616580310881, | |
| "grad_norm": 0.3922266662120819, | |
| "learning_rate": 3.694473329154778e-05, | |
| "loss": 0.7364, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6303972366148531, | |
| "grad_norm": 0.38554486632347107, | |
| "learning_rate": 3.45139266054715e-05, | |
| "loss": 0.7684, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6390328151986183, | |
| "grad_norm": 0.3428330421447754, | |
| "learning_rate": 3.21490588442868e-05, | |
| "loss": 0.7346, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6476683937823834, | |
| "grad_norm": 0.41248151659965515, | |
| "learning_rate": 2.9852511229367865e-05, | |
| "loss": 0.7606, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6563039723661486, | |
| "grad_norm": 0.36212557554244995, | |
| "learning_rate": 2.7626596189492983e-05, | |
| "loss": 0.7511, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6649395509499136, | |
| "grad_norm": 0.36411234736442566, | |
| "learning_rate": 2.5473555032424533e-05, | |
| "loss": 0.7316, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6735751295336787, | |
| "grad_norm": 0.34422898292541504, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.7304, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6822107081174439, | |
| "grad_norm": 0.4062047302722931, | |
| "learning_rate": 2.139469052572127e-05, | |
| "loss": 0.7369, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.690846286701209, | |
| "grad_norm": 0.3605830669403076, | |
| "learning_rate": 1.947297424689414e-05, | |
| "loss": 0.7435, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6994818652849741, | |
| "grad_norm": 0.39452221989631653, | |
| "learning_rate": 1.763234185701673e-05, | |
| "loss": 0.733, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7081174438687392, | |
| "grad_norm": 0.5000078082084656, | |
| "learning_rate": 1.587464671688187e-05, | |
| "loss": 0.7747, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7167530224525043, | |
| "grad_norm": 0.35579192638397217, | |
| "learning_rate": 1.4201658676502294e-05, | |
| "loss": 0.7468, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7253886010362695, | |
| "grad_norm": 0.36388713121414185, | |
| "learning_rate": 1.2615062293021507e-05, | |
| "loss": 0.74, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7340241796200345, | |
| "grad_norm": 0.352222740650177, | |
| "learning_rate": 1.1116455134507664e-05, | |
| "loss": 0.735, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7426597582037997, | |
| "grad_norm": 0.42218002676963806, | |
| "learning_rate": 9.707346171337894e-06, | |
| "loss": 0.7086, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7512953367875648, | |
| "grad_norm": 0.38099607825279236, | |
| "learning_rate": 8.38915425679304e-06, | |
| "loss": 0.7615, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7599309153713298, | |
| "grad_norm": 0.35596323013305664, | |
| "learning_rate": 7.163206698392744e-06, | |
| "loss": 0.7398, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.768566493955095, | |
| "grad_norm": 0.38860464096069336, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.7362, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7772020725388601, | |
| "grad_norm": 0.4052976965904236, | |
| "learning_rate": 4.992888225905468e-06, | |
| "loss": 0.7543, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7858376511226253, | |
| "grad_norm": 0.3824405372142792, | |
| "learning_rate": 4.050702638550275e-06, | |
| "loss": 0.724, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7944732297063903, | |
| "grad_norm": 0.40496277809143066, | |
| "learning_rate": 3.2051298603643753e-06, | |
| "loss": 0.706, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8031088082901554, | |
| "grad_norm": 0.3866259455680847, | |
| "learning_rate": 2.4570213114592954e-06, | |
| "loss": 0.741, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8117443868739206, | |
| "grad_norm": 0.35403546690940857, | |
| "learning_rate": 1.8071302737293295e-06, | |
| "loss": 0.7306, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8203799654576857, | |
| "grad_norm": 0.4546634554862976, | |
| "learning_rate": 1.2561111323605712e-06, | |
| "loss": 0.7597, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8290155440414507, | |
| "grad_norm": 0.3442727029323578, | |
| "learning_rate": 8.04518716920466e-07, | |
| "loss": 0.7845, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8376511226252159, | |
| "grad_norm": 0.37662404775619507, | |
| "learning_rate": 4.5280774269154115e-07, | |
| "loss": 0.7295, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.846286701208981, | |
| "grad_norm": 0.34127819538116455, | |
| "learning_rate": 2.0133235281156736e-07, | |
| "loss": 0.7542, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8549222797927462, | |
| "grad_norm": 0.38656285405158997, | |
| "learning_rate": 5.0345761681491746e-08, | |
| "loss": 0.7498, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8635578583765112, | |
| "grad_norm": 0.3828294575214386, | |
| "learning_rate": 0.0, | |
| "loss": 0.74, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1000, | |
| "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": 2.3420339184869376e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |