| { | |
| "best_metric": 0.24043463170528412, | |
| "best_model_checkpoint": "./ryan_model314/checkpoint-500", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 1252, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.962435245513916, | |
| "learning_rate": 0.00019840255591054313, | |
| "loss": 0.5392, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.411149024963379, | |
| "learning_rate": 0.00019680511182108628, | |
| "loss": 0.3886, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.624254584312439, | |
| "learning_rate": 0.0001952076677316294, | |
| "loss": 0.3602, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 1.0637298822402954, | |
| "learning_rate": 0.00019361022364217253, | |
| "loss": 0.3841, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.1236467361450195, | |
| "learning_rate": 0.00019201277955271565, | |
| "loss": 0.3312, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.6737650632858276, | |
| "learning_rate": 0.0001904153354632588, | |
| "loss": 0.3419, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.0528877973556519, | |
| "learning_rate": 0.00018881789137380192, | |
| "loss": 0.348, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 1.2546306848526, | |
| "learning_rate": 0.00018722044728434505, | |
| "loss": 0.3884, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.552256464958191, | |
| "learning_rate": 0.0001856230031948882, | |
| "loss": 0.4075, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.334892988204956, | |
| "learning_rate": 0.00018402555910543132, | |
| "loss": 0.3682, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.3208509385585785, | |
| "eval_na_accuracy": 0.942, | |
| "eval_ordinal_accuracy": 0.5268065268065268, | |
| "eval_runtime": 54.8874, | |
| "eval_samples_per_second": 9.11, | |
| "eval_steps_per_second": 1.148, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.1707366704940796, | |
| "learning_rate": 0.00018242811501597444, | |
| "loss": 0.274, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.9099497199058533, | |
| "learning_rate": 0.00018083067092651756, | |
| "loss": 0.3335, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.9688892960548401, | |
| "learning_rate": 0.00017923322683706071, | |
| "loss": 0.2867, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.5258199572563171, | |
| "learning_rate": 0.00017763578274760384, | |
| "loss": 0.2557, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.8469595313072205, | |
| "learning_rate": 0.000176038338658147, | |
| "loss": 0.3041, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.846753716468811, | |
| "learning_rate": 0.0001744408945686901, | |
| "loss": 0.2766, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.4423494338989258, | |
| "learning_rate": 0.00017284345047923323, | |
| "loss": 0.2555, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 2.977750062942505, | |
| "learning_rate": 0.00017124600638977638, | |
| "loss": 0.2742, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.8587075471878052, | |
| "learning_rate": 0.00016964856230031948, | |
| "loss": 0.3036, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.7394629120826721, | |
| "learning_rate": 0.00016805111821086263, | |
| "loss": 0.2877, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_loss": 0.2782871127128601, | |
| "eval_na_accuracy": 0.944, | |
| "eval_ordinal_accuracy": 0.5920745920745921, | |
| "eval_runtime": 19.4606, | |
| "eval_samples_per_second": 25.693, | |
| "eval_steps_per_second": 3.237, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.093324899673462, | |
| "learning_rate": 0.00016645367412140575, | |
| "loss": 0.2535, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 1.0213277339935303, | |
| "learning_rate": 0.0001648562300319489, | |
| "loss": 0.2513, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.1321722269058228, | |
| "learning_rate": 0.00016325878594249202, | |
| "loss": 0.3015, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 3.1191928386688232, | |
| "learning_rate": 0.00016166134185303515, | |
| "loss": 0.2307, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.107519507408142, | |
| "learning_rate": 0.0001600638977635783, | |
| "loss": 0.3198, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.818793535232544, | |
| "learning_rate": 0.00015846645367412142, | |
| "loss": 0.2959, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.4491446912288666, | |
| "learning_rate": 0.00015686900958466454, | |
| "loss": 0.212, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.838297963142395, | |
| "learning_rate": 0.00015527156549520767, | |
| "loss": 0.2836, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.4379059076309204, | |
| "learning_rate": 0.00015367412140575082, | |
| "loss": 0.309, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.274575710296631, | |
| "learning_rate": 0.00015207667731629394, | |
| "loss": 0.3079, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_loss": 0.2544197142124176, | |
| "eval_na_accuracy": 0.958, | |
| "eval_ordinal_accuracy": 0.5337995337995338, | |
| "eval_runtime": 19.3178, | |
| "eval_samples_per_second": 25.883, | |
| "eval_steps_per_second": 3.261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.7160611748695374, | |
| "learning_rate": 0.00015047923322683706, | |
| "loss": 0.223, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.5523737668991089, | |
| "learning_rate": 0.0001488817891373802, | |
| "loss": 0.2784, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.5268480777740479, | |
| "learning_rate": 0.00014728434504792333, | |
| "loss": 0.1671, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.7597477436065674, | |
| "learning_rate": 0.00014568690095846646, | |
| "loss": 0.1519, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.7296491861343384, | |
| "learning_rate": 0.00014408945686900958, | |
| "loss": 0.223, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.3579563796520233, | |
| "learning_rate": 0.00014249201277955273, | |
| "loss": 0.1383, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 1.5826038122177124, | |
| "learning_rate": 0.00014089456869009585, | |
| "loss": 0.2555, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 1.290208339691162, | |
| "learning_rate": 0.000139297124600639, | |
| "loss": 0.1482, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.9879806637763977, | |
| "learning_rate": 0.00013769968051118212, | |
| "loss": 0.1714, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6665166616439819, | |
| "learning_rate": 0.00013610223642172525, | |
| "loss": 0.1901, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_loss": 0.26315367221832275, | |
| "eval_na_accuracy": 0.948, | |
| "eval_ordinal_accuracy": 0.6060606060606061, | |
| "eval_runtime": 19.5095, | |
| "eval_samples_per_second": 25.628, | |
| "eval_steps_per_second": 3.229, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.5691338181495667, | |
| "learning_rate": 0.00013450479233226837, | |
| "loss": 0.21, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.76593017578125, | |
| "learning_rate": 0.0001329073482428115, | |
| "loss": 0.1704, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.6972767114639282, | |
| "learning_rate": 0.00013130990415335464, | |
| "loss": 0.1911, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.6530088186264038, | |
| "learning_rate": 0.00012971246006389777, | |
| "loss": 0.128, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.8316710591316223, | |
| "learning_rate": 0.00012811501597444092, | |
| "loss": 0.2046, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.5624408721923828, | |
| "learning_rate": 0.00012651757188498404, | |
| "loss": 0.1888, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.5218725800514221, | |
| "learning_rate": 0.00012492012779552716, | |
| "loss": 0.2096, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.7621105909347534, | |
| "learning_rate": 0.00012332268370607028, | |
| "loss": 0.1682, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.9870132803916931, | |
| "learning_rate": 0.00012172523961661342, | |
| "loss": 0.1606, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5003547072410583, | |
| "learning_rate": 0.00012012779552715656, | |
| "loss": 0.277, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 0.24043463170528412, | |
| "eval_na_accuracy": 0.96, | |
| "eval_ordinal_accuracy": 0.5920745920745921, | |
| "eval_runtime": 19.0104, | |
| "eval_samples_per_second": 26.301, | |
| "eval_steps_per_second": 3.314, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 2.554410457611084, | |
| "learning_rate": 0.00011853035143769968, | |
| "loss": 0.2163, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.577261745929718, | |
| "learning_rate": 0.00011693290734824283, | |
| "loss": 0.1652, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.722175121307373, | |
| "learning_rate": 0.00011533546325878595, | |
| "loss": 0.1061, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.013876438140869, | |
| "learning_rate": 0.00011373801916932908, | |
| "loss": 0.2078, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.5454280376434326, | |
| "learning_rate": 0.00011214057507987221, | |
| "loss": 0.238, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.7590613961219788, | |
| "learning_rate": 0.00011054313099041533, | |
| "loss": 0.1868, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 1.2678501605987549, | |
| "learning_rate": 0.00010894568690095847, | |
| "loss": 0.2365, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.9851756691932678, | |
| "learning_rate": 0.0001073482428115016, | |
| "loss": 0.2362, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.6259474754333496, | |
| "learning_rate": 0.00010575079872204474, | |
| "loss": 0.2368, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.7762842178344727, | |
| "learning_rate": 0.00010415335463258787, | |
| "loss": 0.2081, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_loss": 0.2585134506225586, | |
| "eval_na_accuracy": 0.95, | |
| "eval_ordinal_accuracy": 0.627039627039627, | |
| "eval_runtime": 19.183, | |
| "eval_samples_per_second": 26.065, | |
| "eval_steps_per_second": 3.284, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.8312740325927734, | |
| "learning_rate": 0.000102555910543131, | |
| "loss": 0.1747, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.1986323595046997, | |
| "learning_rate": 0.00010095846645367413, | |
| "loss": 0.146, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.6852553486824036, | |
| "learning_rate": 9.936102236421726e-05, | |
| "loss": 0.1098, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.49778783321380615, | |
| "learning_rate": 9.77635782747604e-05, | |
| "loss": 0.0807, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.43836385011672974, | |
| "learning_rate": 9.616613418530351e-05, | |
| "loss": 0.0785, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.4496062695980072, | |
| "learning_rate": 9.456869009584664e-05, | |
| "loss": 0.1007, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.6552535891532898, | |
| "learning_rate": 9.297124600638978e-05, | |
| "loss": 0.0837, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.8565073013305664, | |
| "learning_rate": 9.137380191693292e-05, | |
| "loss": 0.1288, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.3386971056461334, | |
| "learning_rate": 8.977635782747604e-05, | |
| "loss": 0.0671, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.6647264957427979, | |
| "learning_rate": 8.817891373801918e-05, | |
| "loss": 0.0809, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_loss": 0.26309889554977417, | |
| "eval_na_accuracy": 0.954, | |
| "eval_ordinal_accuracy": 0.6736596736596736, | |
| "eval_runtime": 18.6607, | |
| "eval_samples_per_second": 26.794, | |
| "eval_steps_per_second": 3.376, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.6069409251213074, | |
| "learning_rate": 8.658146964856231e-05, | |
| "loss": 0.0765, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.3801295757293701, | |
| "learning_rate": 8.498402555910544e-05, | |
| "loss": 0.1077, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.7153878808021545, | |
| "learning_rate": 8.338658146964856e-05, | |
| "loss": 0.075, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.727155327796936, | |
| "learning_rate": 8.17891373801917e-05, | |
| "loss": 0.0639, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.7251117825508118, | |
| "learning_rate": 8.019169329073483e-05, | |
| "loss": 0.1146, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.7786515951156616, | |
| "learning_rate": 7.859424920127795e-05, | |
| "loss": 0.1122, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.6191527843475342, | |
| "learning_rate": 7.699680511182109e-05, | |
| "loss": 0.0867, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.4263085722923279, | |
| "learning_rate": 7.539936102236423e-05, | |
| "loss": 0.1325, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 1.6009018421173096, | |
| "learning_rate": 7.380191693290735e-05, | |
| "loss": 0.0693, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.37769052386283875, | |
| "learning_rate": 7.220447284345049e-05, | |
| "loss": 0.1, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_loss": 0.2692907154560089, | |
| "eval_na_accuracy": 0.958, | |
| "eval_ordinal_accuracy": 0.6363636363636364, | |
| "eval_runtime": 19.3788, | |
| "eval_samples_per_second": 25.801, | |
| "eval_steps_per_second": 3.251, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.45006510615348816, | |
| "learning_rate": 7.060702875399361e-05, | |
| "loss": 0.0645, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.813955545425415, | |
| "learning_rate": 6.900958466453674e-05, | |
| "loss": 0.0583, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.7967355847358704, | |
| "learning_rate": 6.741214057507987e-05, | |
| "loss": 0.1065, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.8020057678222656, | |
| "learning_rate": 6.5814696485623e-05, | |
| "loss": 0.0821, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.7547793388366699, | |
| "learning_rate": 6.421725239616614e-05, | |
| "loss": 0.0983, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.42487284541130066, | |
| "learning_rate": 6.261980830670928e-05, | |
| "loss": 0.1221, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.3974541425704956, | |
| "learning_rate": 6.1022364217252406e-05, | |
| "loss": 0.108, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.42565950751304626, | |
| "learning_rate": 5.942492012779552e-05, | |
| "loss": 0.106, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.5634491443634033, | |
| "learning_rate": 5.782747603833866e-05, | |
| "loss": 0.0765, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.7806987166404724, | |
| "learning_rate": 5.623003194888179e-05, | |
| "loss": 0.0924, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_loss": 0.2566128671169281, | |
| "eval_na_accuracy": 0.962, | |
| "eval_ordinal_accuracy": 0.6503496503496503, | |
| "eval_runtime": 19.1674, | |
| "eval_samples_per_second": 26.086, | |
| "eval_steps_per_second": 3.287, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 1.5311517715454102, | |
| "learning_rate": 5.4632587859424925e-05, | |
| "loss": 0.0779, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.8630849123001099, | |
| "learning_rate": 5.3035143769968054e-05, | |
| "loss": 0.0697, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.39383065700531006, | |
| "learning_rate": 5.1437699680511184e-05, | |
| "loss": 0.065, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.45358964800834656, | |
| "learning_rate": 4.984025559105431e-05, | |
| "loss": 0.0489, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.39319777488708496, | |
| "learning_rate": 4.824281150159744e-05, | |
| "loss": 0.0311, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.6434493064880371, | |
| "learning_rate": 4.664536741214058e-05, | |
| "loss": 0.0424, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.4773056209087372, | |
| "learning_rate": 4.504792332268371e-05, | |
| "loss": 0.0326, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.48347094655036926, | |
| "learning_rate": 4.345047923322684e-05, | |
| "loss": 0.0312, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.410236120223999, | |
| "learning_rate": 4.185303514376997e-05, | |
| "loss": 0.0283, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 0.3112216889858246, | |
| "learning_rate": 4.0255591054313104e-05, | |
| "loss": 0.0304, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "eval_loss": 0.2782021760940552, | |
| "eval_na_accuracy": 0.954, | |
| "eval_ordinal_accuracy": 0.6526806526806527, | |
| "eval_runtime": 18.9723, | |
| "eval_samples_per_second": 26.354, | |
| "eval_steps_per_second": 3.321, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.3799457550048828, | |
| "learning_rate": 3.8658146964856234e-05, | |
| "loss": 0.0622, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 0.3940556049346924, | |
| "learning_rate": 3.7060702875399364e-05, | |
| "loss": 0.0275, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.5390946269035339, | |
| "learning_rate": 3.546325878594249e-05, | |
| "loss": 0.0531, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 0.5048585534095764, | |
| "learning_rate": 3.386581469648562e-05, | |
| "loss": 0.03, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.4280671775341034, | |
| "learning_rate": 3.242811501597444e-05, | |
| "loss": 0.0474, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.34902918338775635, | |
| "learning_rate": 3.083067092651757e-05, | |
| "loss": 0.0318, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 0.4760427176952362, | |
| "learning_rate": 2.9233226837060707e-05, | |
| "loss": 0.0285, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.15729285776615143, | |
| "learning_rate": 2.7635782747603834e-05, | |
| "loss": 0.0496, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 0.38401302695274353, | |
| "learning_rate": 2.6038338658146967e-05, | |
| "loss": 0.0252, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.3598209023475647, | |
| "learning_rate": 2.44408945686901e-05, | |
| "loss": 0.0256, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "eval_loss": 0.26621493697166443, | |
| "eval_na_accuracy": 0.96, | |
| "eval_ordinal_accuracy": 0.6643356643356644, | |
| "eval_runtime": 19.2883, | |
| "eval_samples_per_second": 25.922, | |
| "eval_steps_per_second": 3.266, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.4313770830631256, | |
| "learning_rate": 2.284345047923323e-05, | |
| "loss": 0.0269, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 0.4230475425720215, | |
| "learning_rate": 2.124600638977636e-05, | |
| "loss": 0.0304, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 0.2662275433540344, | |
| "learning_rate": 1.964856230031949e-05, | |
| "loss": 0.0176, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.791589617729187, | |
| "learning_rate": 1.805111821086262e-05, | |
| "loss": 0.0307, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.49746203422546387, | |
| "learning_rate": 1.645367412140575e-05, | |
| "loss": 0.0217, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.18784604966640472, | |
| "learning_rate": 1.485623003194888e-05, | |
| "loss": 0.0206, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 0.47459328174591064, | |
| "learning_rate": 1.3258785942492014e-05, | |
| "loss": 0.0265, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 0.2313966453075409, | |
| "learning_rate": 1.1661341853035145e-05, | |
| "loss": 0.0313, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.4523174464702606, | |
| "learning_rate": 1.0063897763578276e-05, | |
| "loss": 0.0622, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 0.17250552773475647, | |
| "learning_rate": 8.466453674121406e-06, | |
| "loss": 0.0424, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "eval_loss": 0.2670270502567291, | |
| "eval_na_accuracy": 0.962, | |
| "eval_ordinal_accuracy": 0.675990675990676, | |
| "eval_runtime": 18.6815, | |
| "eval_samples_per_second": 26.764, | |
| "eval_steps_per_second": 3.372, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.3159545660018921, | |
| "learning_rate": 6.869009584664538e-06, | |
| "loss": 0.0209, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.5657308101654053, | |
| "learning_rate": 5.2715654952076674e-06, | |
| "loss": 0.019, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 0.4505397379398346, | |
| "learning_rate": 3.6741214057507987e-06, | |
| "loss": 0.0249, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.21818359196186066, | |
| "learning_rate": 2.0766773162939296e-06, | |
| "loss": 0.0226, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 0.28112998604774475, | |
| "learning_rate": 4.792332268370607e-07, | |
| "loss": 0.027, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 1252, | |
| "total_flos": 1.5498953551872e+18, | |
| "train_loss": 0.15608444792060808, | |
| "train_runtime": 1745.0942, | |
| "train_samples_per_second": 11.461, | |
| "train_steps_per_second": 0.717 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1252, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "total_flos": 1.5498953551872e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |