| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 100, | |
| "global_step": 1315, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 4.0682622539611835, | |
| "learning_rate": 5e-06, | |
| "loss": 4.2205, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.312709224854891, | |
| "learning_rate": 1e-05, | |
| "loss": 4.2858, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 3.6601980442252438, | |
| "learning_rate": 1.5e-05, | |
| "loss": 4.1264, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.812299442715205, | |
| "learning_rate": 2e-05, | |
| "loss": 4.302, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.116526134864302, | |
| "learning_rate": 2.5e-05, | |
| "loss": 4.1039, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.6960273710433897, | |
| "learning_rate": 3e-05, | |
| "loss": 3.7882, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.511832693580313, | |
| "learning_rate": 3.5e-05, | |
| "loss": 3.4971, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.9433322085763716, | |
| "learning_rate": 4e-05, | |
| "loss": 3.1992, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.6894476126655664, | |
| "learning_rate": 4.5e-05, | |
| "loss": 3.0814, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.073675333188434, | |
| "learning_rate": 5e-05, | |
| "loss": 2.8756, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 1.4627544436421924, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.7048, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.99973700742081, | |
| "learning_rate": 6e-05, | |
| "loss": 2.6456, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.6172270630421473, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.6635, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.1295069230502908, | |
| "learning_rate": 7e-05, | |
| "loss": 2.59, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.2649325235735724, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.44, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 1.1789342809576489, | |
| "learning_rate": 8e-05, | |
| "loss": 2.5202, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.7081838083903702, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.367, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.7108144063965985, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4458, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.6609440232385762, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.3898, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.5625207480279918, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3829, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.6335176424475211, | |
| "learning_rate": 0.000105, | |
| "loss": 2.3701, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.5478182407846386, | |
| "learning_rate": 0.00011000000000000002, | |
| "loss": 2.3466, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.5542241828994607, | |
| "learning_rate": 0.00011499999999999999, | |
| "loss": 2.3094, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.5391779694945925, | |
| "learning_rate": 0.00012, | |
| "loss": 2.2829, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.5560267338571905, | |
| "learning_rate": 0.000125, | |
| "loss": 2.3031, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.5899065231096706, | |
| "learning_rate": 0.00013000000000000002, | |
| "loss": 2.3218, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.5822360971814909, | |
| "learning_rate": 0.00013500000000000003, | |
| "loss": 2.2678, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.5713806277891299, | |
| "learning_rate": 0.00014, | |
| "loss": 2.2085, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.5778957318063205, | |
| "learning_rate": 0.000145, | |
| "loss": 2.1972, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.6239096985024343, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 2.157, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.664987114969562, | |
| "learning_rate": 0.000155, | |
| "loss": 2.144, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.6602627781365719, | |
| "learning_rate": 0.00016, | |
| "loss": 2.1819, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.6953248375158279, | |
| "learning_rate": 0.000165, | |
| "loss": 2.0508, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.750259033408061, | |
| "learning_rate": 0.00017, | |
| "loss": 2.129, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.6643380168708766, | |
| "learning_rate": 0.000175, | |
| "loss": 2.014, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.8486382402556593, | |
| "learning_rate": 0.00018, | |
| "loss": 1.9433, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7414001872340891, | |
| "learning_rate": 0.00018500000000000002, | |
| "loss": 2.0308, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.7295050222316729, | |
| "learning_rate": 0.00019, | |
| "loss": 1.9284, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.7235284135505695, | |
| "learning_rate": 0.000195, | |
| "loss": 1.8847, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.753068728806512, | |
| "learning_rate": 0.0002, | |
| "loss": 1.8865, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7560315372185343, | |
| "learning_rate": 0.00019999969643677332, | |
| "loss": 1.9206, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7536264695992508, | |
| "learning_rate": 0.00019999878574893627, | |
| "loss": 1.8679, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.7688437635155836, | |
| "learning_rate": 0.0001999972679420179, | |
| "loss": 1.8956, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7269625532226869, | |
| "learning_rate": 0.0001999951430252332, | |
| "loss": 1.7802, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.7518593665603851, | |
| "learning_rate": 0.00019999241101148306, | |
| "loss": 1.8219, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8030012654181582, | |
| "learning_rate": 0.00019998907191735434, | |
| "loss": 1.7884, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.7431537767433151, | |
| "learning_rate": 0.00019998512576311953, | |
| "loss": 1.7157, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.8789668942389129, | |
| "learning_rate": 0.00019998057257273675, | |
| "loss": 1.705, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.8821830829681667, | |
| "learning_rate": 0.00019997541237384966, | |
| "loss": 1.7197, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.7827120449376896, | |
| "learning_rate": 0.0001999696451977872, | |
| "loss": 1.744, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.7824589333224853, | |
| "learning_rate": 0.00019996327107956333, | |
| "loss": 1.5894, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.805173848765375, | |
| "learning_rate": 0.00019995629005787713, | |
| "loss": 1.764, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.8408625701287021, | |
| "learning_rate": 0.00019994870217511217, | |
| "loss": 1.7377, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.7546347988189869, | |
| "learning_rate": 0.0001999405074773365, | |
| "loss": 1.7158, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.821101064412719, | |
| "learning_rate": 0.0001999317060143023, | |
| "loss": 1.5648, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.7901726402638417, | |
| "learning_rate": 0.00019992229783944557, | |
| "loss": 1.5496, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.8488465994765656, | |
| "learning_rate": 0.00019991228300988585, | |
| "loss": 1.5493, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.8859226699933315, | |
| "learning_rate": 0.0001999016615864258, | |
| "loss": 1.4703, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.9572611655447181, | |
| "learning_rate": 0.0001998904336335509, | |
| "loss": 1.5522, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.8654728304283881, | |
| "learning_rate": 0.00019987859921942903, | |
| "loss": 1.4136, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.7903035834496105, | |
| "learning_rate": 0.00019986615841591002, | |
| "loss": 1.5446, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.8146733074382629, | |
| "learning_rate": 0.0001998531112985253, | |
| "loss": 1.4741, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.8150030052550084, | |
| "learning_rate": 0.00019983945794648734, | |
| "loss": 1.4877, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.8980109486617155, | |
| "learning_rate": 0.00019982519844268933, | |
| "loss": 1.4981, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.9375060640657445, | |
| "learning_rate": 0.00019981033287370443, | |
| "loss": 1.5115, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.8164746637962466, | |
| "learning_rate": 0.00019979486132978545, | |
| "loss": 1.4431, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.8458741113281569, | |
| "learning_rate": 0.0001997787839048642, | |
| "loss": 1.3408, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.8444471080538805, | |
| "learning_rate": 0.00019976210069655104, | |
| "loss": 1.3728, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.9081815525327502, | |
| "learning_rate": 0.0001997448118061341, | |
| "loss": 1.3964, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.9703756561572067, | |
| "learning_rate": 0.00019972691733857883, | |
| "loss": 1.2744, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.8300846731216797, | |
| "learning_rate": 0.00019970841740252725, | |
| "loss": 1.201, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.8970681814115683, | |
| "learning_rate": 0.00019968931211029734, | |
| "loss": 1.3049, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.9203036618648276, | |
| "learning_rate": 0.00019966960157788248, | |
| "loss": 1.3423, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.8554417095674692, | |
| "learning_rate": 0.00019964928592495045, | |
| "loss": 1.2076, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.8693522008551818, | |
| "learning_rate": 0.00019962836527484296, | |
| "loss": 1.2388, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.8955501483204824, | |
| "learning_rate": 0.0001996068397545748, | |
| "loss": 1.2769, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.8972853945681226, | |
| "learning_rate": 0.00019958470949483318, | |
| "loss": 1.2043, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.9200733943459781, | |
| "learning_rate": 0.00019956197462997667, | |
| "loss": 1.1127, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.046096681591468, | |
| "learning_rate": 0.00019953863529803466, | |
| "loss": 1.1452, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.0188934085587635, | |
| "learning_rate": 0.00019951469164070646, | |
| "loss": 1.1164, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.8178106254313157, | |
| "learning_rate": 0.00019949014380336028, | |
| "loss": 1.0532, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.7925775197670102, | |
| "learning_rate": 0.00019946499193503262, | |
| "loss": 1.0213, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8643770007688552, | |
| "learning_rate": 0.000199439236188427, | |
| "loss": 1.204, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8294076531102257, | |
| "learning_rate": 0.0001994128767199135, | |
| "loss": 1.1476, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8800850098334678, | |
| "learning_rate": 0.0001993859136895274, | |
| "loss": 1.0514, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.9129898809803545, | |
| "learning_rate": 0.0001993583472609683, | |
| "loss": 1.0532, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.9139962309722894, | |
| "learning_rate": 0.00019933017760159937, | |
| "loss": 1.116, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.8049163381340136, | |
| "learning_rate": 0.00019930140488244602, | |
| "loss": 1.0375, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.7506103786806886, | |
| "learning_rate": 0.0001992720292781951, | |
| "loss": 0.9954, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.8494376985530913, | |
| "learning_rate": 0.0001992420509671936, | |
| "loss": 1.011, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.0838465138358377, | |
| "learning_rate": 0.0001992114701314478, | |
| "loss": 1.1547, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.7780322978386432, | |
| "learning_rate": 0.00019918028695662207, | |
| "loss": 0.8977, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.7445309005650187, | |
| "learning_rate": 0.00019914850163203768, | |
| "loss": 0.9499, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.75618752117177, | |
| "learning_rate": 0.00019911611435067172, | |
| "loss": 0.9058, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.922514343070264, | |
| "learning_rate": 0.00019908312530915603, | |
| "loss": 0.9812, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.8335423463861146, | |
| "learning_rate": 0.00019904953470777575, | |
| "loss": 0.9237, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.8569297364051901, | |
| "learning_rate": 0.0001990153427504683, | |
| "loss": 0.9101, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.8431577631707102, | |
| "learning_rate": 0.00019898054964482214, | |
| "loss": 0.8826, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.8393428609756466, | |
| "learning_rate": 0.00019894515560207537, | |
| "loss": 0.9449, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.8009262477457554, | |
| "learning_rate": 0.0001989091608371146, | |
| "loss": 0.8172, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_blimp_filtered_avg": 0.7153731343283583, | |
| "eval_blimp_filtered_std": 0.004995278040149823, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_blimp_supplement_avg": 0.8275862068965517, | |
| "eval_blimp_supplement_std": 0.016574057417324883, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_vqa_filtered_avg": 0.5, | |
| "eval_vqa_filtered_std": 0.050251890762960605, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_winoground_filtered_avg": 0.64, | |
| "eval_winoground_filtered_std": 0.048241815132442176, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.7731939394783305, | |
| "learning_rate": 0.0001988725655684736, | |
| "loss": 0.841, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.9253945491288401, | |
| "learning_rate": 0.00019883537001833188, | |
| "loss": 0.9227, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.9312582898914622, | |
| "learning_rate": 0.0001987975744125135, | |
| "loss": 0.9283, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.781149897014019, | |
| "learning_rate": 0.00019875917898048558, | |
| "loss": 0.7902, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.8132523662925432, | |
| "learning_rate": 0.0001987201839553569, | |
| "loss": 0.8595, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.8368435917049845, | |
| "learning_rate": 0.00019868058957387663, | |
| "loss": 0.8643, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.7460582173754553, | |
| "learning_rate": 0.00019864039607643273, | |
| "loss": 0.7484, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.7789594633508866, | |
| "learning_rate": 0.0001985996037070505, | |
| "loss": 0.7889, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.8428485721906823, | |
| "learning_rate": 0.00019855821271339125, | |
| "loss": 0.838, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.8694838248366092, | |
| "learning_rate": 0.00019851622334675066, | |
| "loss": 0.77, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.8994640080800599, | |
| "learning_rate": 0.00019847363586205727, | |
| "loss": 0.7702, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.9931271763497148, | |
| "learning_rate": 0.00019843045051787096, | |
| "loss": 0.8737, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.8186757528083024, | |
| "learning_rate": 0.00019838666757638135, | |
| "loss": 0.7276, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.7976246474968561, | |
| "learning_rate": 0.0001983422873034063, | |
| "loss": 0.7967, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.7725738117793295, | |
| "learning_rate": 0.0001982973099683902, | |
| "loss": 0.8214, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.7497131152895721, | |
| "learning_rate": 0.00019825173584440232, | |
| "loss": 0.7069, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.8096175366747367, | |
| "learning_rate": 0.0001982055652081352, | |
| "loss": 0.699, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.8131636775412501, | |
| "learning_rate": 0.00019815879833990304, | |
| "loss": 0.7479, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.8103259093372099, | |
| "learning_rate": 0.00019811143552363983, | |
| "loss": 0.7013, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.8280999953040724, | |
| "learning_rate": 0.00019806347704689778, | |
| "loss": 0.6887, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.8116880398114632, | |
| "learning_rate": 0.00019801492320084546, | |
| "loss": 0.7934, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.8303791216043742, | |
| "learning_rate": 0.00019796577428026616, | |
| "loss": 0.8011, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.7619123156871999, | |
| "learning_rate": 0.00019791603058355595, | |
| "loss": 0.7217, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.7733520932459683, | |
| "learning_rate": 0.00019786569241272197, | |
| "loss": 0.6981, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.7846699948014345, | |
| "learning_rate": 0.00019781476007338058, | |
| "loss": 0.7401, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.7963675285315092, | |
| "learning_rate": 0.00019776323387475547, | |
| "loss": 0.705, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.7856922504610907, | |
| "learning_rate": 0.00019771111412967583, | |
| "loss": 0.6999, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.831698101406348, | |
| "learning_rate": 0.0001976584011545744, | |
| "loss": 0.7021, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.8387219512744284, | |
| "learning_rate": 0.00019760509526948566, | |
| "loss": 0.7296, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.8945842296083665, | |
| "learning_rate": 0.00019755119679804367, | |
| "loss": 0.7529, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.7498731306068172, | |
| "learning_rate": 0.00019749670606748033, | |
| "loss": 0.6911, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.7247722560830141, | |
| "learning_rate": 0.0001974416234086233, | |
| "loss": 0.6402, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.7627410009794175, | |
| "learning_rate": 0.00019738594915589397, | |
| "loss": 0.6702, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.7772008131178036, | |
| "learning_rate": 0.00019732968364730545, | |
| "loss": 0.6963, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.8275796285376315, | |
| "learning_rate": 0.00019727282722446047, | |
| "loss": 0.5929, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.7831120443420706, | |
| "learning_rate": 0.0001972153802325495, | |
| "loss": 0.6179, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.8292277458348148, | |
| "learning_rate": 0.0001971573430203484, | |
| "loss": 0.6375, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.8340476959630893, | |
| "learning_rate": 0.00019709871594021642, | |
| "loss": 0.6766, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.7994798406256604, | |
| "learning_rate": 0.00019703949934809408, | |
| "loss": 0.6239, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.7425672818805296, | |
| "learning_rate": 0.00019697969360350098, | |
| "loss": 0.6334, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.7443030221681443, | |
| "learning_rate": 0.00019691929906953356, | |
| "loss": 0.6278, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.7858836231557178, | |
| "learning_rate": 0.0001968583161128631, | |
| "loss": 0.6494, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.7748084569986792, | |
| "learning_rate": 0.00019679674510373325, | |
| "loss": 0.6408, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.7352594538958535, | |
| "learning_rate": 0.00019673458641595784, | |
| "loss": 0.54, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.9565541632170729, | |
| "learning_rate": 0.00019667184042691875, | |
| "loss": 0.6669, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.8159705350255112, | |
| "learning_rate": 0.00019660850751756348, | |
| "loss": 0.549, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.853659410491719, | |
| "learning_rate": 0.00019654458807240283, | |
| "loss": 0.703, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.7233148520384484, | |
| "learning_rate": 0.0001964800824795087, | |
| "loss": 0.5607, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.8017084720492917, | |
| "learning_rate": 0.00019641499113051157, | |
| "loss": 0.6522, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.6987528025960802, | |
| "learning_rate": 0.00019634931442059832, | |
| "loss": 0.5149, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.8039264184533087, | |
| "learning_rate": 0.00019628305274850956, | |
| "loss": 0.7189, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.795017894232834, | |
| "learning_rate": 0.00019621620651653744, | |
| "loss": 0.6096, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.7625532547573897, | |
| "learning_rate": 0.00019614877613052312, | |
| "loss": 0.583, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.8479800597198418, | |
| "learning_rate": 0.00019608076199985433, | |
| "loss": 0.5841, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.7984666488345297, | |
| "learning_rate": 0.00019601216453746283, | |
| "loss": 0.617, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.7291592602065788, | |
| "learning_rate": 0.00019594298415982194, | |
| "loss": 0.5409, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.7444160049758751, | |
| "learning_rate": 0.0001958732212869441, | |
| "loss": 0.5235, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.7698434281876576, | |
| "learning_rate": 0.00019580287634237808, | |
| "loss": 0.518, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.658666402238984, | |
| "learning_rate": 0.00019573194975320673, | |
| "loss": 0.4665, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.9088095098968566, | |
| "learning_rate": 0.0001956604419500441, | |
| "loss": 0.719, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.6790919494296814, | |
| "learning_rate": 0.00019558835336703294, | |
| "loss": 0.5317, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.7461814218746576, | |
| "learning_rate": 0.00019551568444184215, | |
| "loss": 0.5383, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.750432430440515, | |
| "learning_rate": 0.00019544243561566403, | |
| "loss": 0.5389, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.7472033806009103, | |
| "learning_rate": 0.00019536860733321152, | |
| "loss": 0.5066, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.7137901167042525, | |
| "learning_rate": 0.00019529420004271567, | |
| "loss": 0.4392, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.8491528440848178, | |
| "learning_rate": 0.00019521921419592283, | |
| "loss": 0.4803, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.888429079441991, | |
| "learning_rate": 0.0001951436502480919, | |
| "loss": 0.4945, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.7568924214482684, | |
| "learning_rate": 0.00019506750865799162, | |
| "loss": 0.5297, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.772439312971248, | |
| "learning_rate": 0.0001949907898878977, | |
| "loss": 0.5506, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.7127854034612457, | |
| "learning_rate": 0.00019491349440359015, | |
| "loss": 0.5152, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.6775965473007823, | |
| "learning_rate": 0.00019483562267435018, | |
| "loss": 0.4436, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.7085484218846827, | |
| "learning_rate": 0.00019475717517295778, | |
| "loss": 0.4726, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.6960992654258619, | |
| "learning_rate": 0.00019467815237568842, | |
| "loss": 0.4145, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.7739939422897306, | |
| "learning_rate": 0.00019459855476231043, | |
| "loss": 0.5368, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.7248885499578243, | |
| "learning_rate": 0.00019451838281608197, | |
| "loss": 0.4216, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.624271177758561, | |
| "learning_rate": 0.00019443763702374812, | |
| "loss": 0.3793, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.6547069254648431, | |
| "learning_rate": 0.00019435631787553795, | |
| "loss": 0.3958, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.7974662713630617, | |
| "learning_rate": 0.00019427442586516155, | |
| "loss": 0.4547, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.6979645264553271, | |
| "learning_rate": 0.00019419196148980693, | |
| "loss": 0.4269, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.7406062144284534, | |
| "learning_rate": 0.0001941089252501372, | |
| "loss": 0.4589, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.7721441112785269, | |
| "learning_rate": 0.00019402531765028722, | |
| "loss": 0.5091, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.6823443706634189, | |
| "learning_rate": 0.00019394113919786094, | |
| "loss": 0.4578, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.6296709496710926, | |
| "learning_rate": 0.00019385639040392803, | |
| "loss": 0.4192, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.7391260784858245, | |
| "learning_rate": 0.00019377107178302074, | |
| "loss": 0.5101, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.7622753386981794, | |
| "learning_rate": 0.00019368518385313107, | |
| "loss": 0.5035, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.7760535426376107, | |
| "learning_rate": 0.00019359872713570732, | |
| "loss": 0.4588, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.7543376381267045, | |
| "learning_rate": 0.00019351170215565114, | |
| "loss": 0.4783, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.761968062534956, | |
| "learning_rate": 0.00019342410944131415, | |
| "loss": 0.4452, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6064844487399527, | |
| "learning_rate": 0.00019333594952449488, | |
| "loss": 0.3948, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.663418856322293, | |
| "learning_rate": 0.00019324722294043558, | |
| "loss": 0.3929, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.7628743041097833, | |
| "learning_rate": 0.00019315793022781877, | |
| "loss": 0.4686, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.7308155097864294, | |
| "learning_rate": 0.00019306807192876412, | |
| "loss": 0.4433, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.6266943285552069, | |
| "learning_rate": 0.00019297764858882514, | |
| "loss": 0.3778, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.7304858297996167, | |
| "learning_rate": 0.00019288666075698588, | |
| "loss": 0.4559, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.793956243255512, | |
| "learning_rate": 0.0001927951089856575, | |
| "loss": 0.5042, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.7521517748378912, | |
| "learning_rate": 0.00019270299383067498, | |
| "loss": 0.4876, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.6541305777620361, | |
| "learning_rate": 0.00019261031585129386, | |
| "loss": 0.4332, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.7459198000292951, | |
| "learning_rate": 0.0001925170756101867, | |
| "loss": 0.4742, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.7351417487869903, | |
| "learning_rate": 0.0001924232736734396, | |
| "loss": 0.4613, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.6880337228260249, | |
| "learning_rate": 0.00019232891061054895, | |
| "loss": 0.4507, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_blimp_filtered_avg": 0.7174626865671642, | |
| "eval_blimp_filtered_std": 0.004966500893623926, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_blimp_supplement_avg": 0.8448275862068966, | |
| "eval_blimp_supplement_std": 0.015878951045947127, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_vqa_filtered_avg": 0.49, | |
| "eval_vqa_filtered_std": 0.05024183937956912, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_winoground_filtered_avg": 0.68, | |
| "eval_winoground_filtered_std": 0.046882617226215034, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.6676123005810946, | |
| "learning_rate": 0.00019223398699441785, | |
| "loss": 0.4456, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.6441892256930322, | |
| "learning_rate": 0.00019213850340135276, | |
| "loss": 0.3842, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.7664915489965373, | |
| "learning_rate": 0.00019204246041105974, | |
| "loss": 0.4655, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.7694952589016031, | |
| "learning_rate": 0.0001919458586066412, | |
| "loss": 0.4042, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.8094634626332351, | |
| "learning_rate": 0.00019184869857459232, | |
| "loss": 0.4796, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.7250896027775108, | |
| "learning_rate": 0.00019175098090479727, | |
| "loss": 0.4493, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.695912278074156, | |
| "learning_rate": 0.00019165270619052595, | |
| "loss": 0.3892, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.6437397715709255, | |
| "learning_rate": 0.00019155387502843013, | |
| "loss": 0.3551, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.6624223376721595, | |
| "learning_rate": 0.00019145448801853989, | |
| "loss": 0.3824, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.7132818106598392, | |
| "learning_rate": 0.0001913545457642601, | |
| "loss": 0.443, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.5835412605472261, | |
| "learning_rate": 0.00019125404887236663, | |
| "loss": 0.2834, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.7082542269732592, | |
| "learning_rate": 0.00019115299795300267, | |
| "loss": 0.3691, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.8422813525765561, | |
| "learning_rate": 0.00019105139361967507, | |
| "loss": 0.4764, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.827672240696612, | |
| "learning_rate": 0.00019094923648925067, | |
| "loss": 0.4577, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.6311700380194617, | |
| "learning_rate": 0.00019084652718195238, | |
| "loss": 0.3641, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.6733244279961911, | |
| "learning_rate": 0.00019074326632135562, | |
| "loss": 0.3914, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.688661851686969, | |
| "learning_rate": 0.00019063945453438432, | |
| "loss": 0.4429, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.593793920110184, | |
| "learning_rate": 0.0001905350924513074, | |
| "loss": 0.3195, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.6401900884414997, | |
| "learning_rate": 0.0001904301807057346, | |
| "loss": 0.3838, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.6222158517510352, | |
| "learning_rate": 0.0001903247199346129, | |
| "loss": 0.3656, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.6346751167318555, | |
| "learning_rate": 0.00019021871077822255, | |
| "loss": 0.3423, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.6573146165491399, | |
| "learning_rate": 0.00019011215388017316, | |
| "loss": 0.3611, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.7506685575783819, | |
| "learning_rate": 0.00019000504988739986, | |
| "loss": 0.4136, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.7933264473843609, | |
| "learning_rate": 0.00018989739945015933, | |
| "loss": 0.4243, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.7056764465112211, | |
| "learning_rate": 0.00018978920322202582, | |
| "loss": 0.393, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.6440690477957951, | |
| "learning_rate": 0.00018968046185988732, | |
| "loss": 0.3339, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.7240338095005376, | |
| "learning_rate": 0.0001895711760239413, | |
| "loss": 0.4307, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.6482583513704759, | |
| "learning_rate": 0.00018946134637769105, | |
| "loss": 0.348, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.6314129612332641, | |
| "learning_rate": 0.00018935097358794144, | |
| "loss": 0.3594, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.6952900742432684, | |
| "learning_rate": 0.00018924005832479478, | |
| "loss": 0.4159, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.6973683572083686, | |
| "learning_rate": 0.00018912860126164707, | |
| "loss": 0.372, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.6228752985295244, | |
| "learning_rate": 0.00018901660307518354, | |
| "loss": 0.3182, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.6853492125246201, | |
| "learning_rate": 0.00018890406444537486, | |
| "loss": 0.3377, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.7047592541956417, | |
| "learning_rate": 0.0001887909860554728, | |
| "loss": 0.3938, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.7514455442163654, | |
| "learning_rate": 0.0001886773685920062, | |
| "loss": 0.4148, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.6032514516825037, | |
| "learning_rate": 0.00018856321274477673, | |
| "loss": 0.2853, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.6566649001522608, | |
| "learning_rate": 0.0001884485192068547, | |
| "loss": 0.333, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.5910387459158075, | |
| "learning_rate": 0.00018833328867457497, | |
| "loss": 0.2821, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.6135310255057351, | |
| "learning_rate": 0.00018821752184753252, | |
| "loss": 0.3528, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.6456839429891811, | |
| "learning_rate": 0.00018810121942857845, | |
| "loss": 0.3332, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.7053660522639559, | |
| "learning_rate": 0.0001879843821238155, | |
| "loss": 0.38, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.6185890160689334, | |
| "learning_rate": 0.00018786701064259383, | |
| "loss": 0.2801, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.7139325318031294, | |
| "learning_rate": 0.00018774910569750673, | |
| "loss": 0.3278, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.7529151224886904, | |
| "learning_rate": 0.00018763066800438636, | |
| "loss": 0.3163, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.7718980285573507, | |
| "learning_rate": 0.00018751169828229927, | |
| "loss": 0.3497, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.6881945450213135, | |
| "learning_rate": 0.00018739219725354212, | |
| "loss": 0.3393, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.6753691180440268, | |
| "learning_rate": 0.00018727216564363723, | |
| "loss": 0.3151, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.6015854239315479, | |
| "learning_rate": 0.00018715160418132832, | |
| "loss": 0.3239, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.5509912220014497, | |
| "learning_rate": 0.00018703051359857586, | |
| "loss": 0.2652, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.6581124878526092, | |
| "learning_rate": 0.00018690889463055283, | |
| "loss": 0.3287, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.6346317349050878, | |
| "learning_rate": 0.0001867867480156402, | |
| "loss": 0.2897, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.658112039759539, | |
| "learning_rate": 0.00018666407449542232, | |
| "loss": 0.2801, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.7964252151066129, | |
| "learning_rate": 0.0001865408748146826, | |
| "loss": 0.3569, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.7087027514563875, | |
| "learning_rate": 0.0001864171497213989, | |
| "loss": 0.319, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.5693751866666873, | |
| "learning_rate": 0.00018629289996673897, | |
| "loss": 0.296, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.4858712761873739, | |
| "learning_rate": 0.00018616812630505597, | |
| "loss": 0.2337, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.6605776129831655, | |
| "learning_rate": 0.0001860428294938838, | |
| "loss": 0.4049, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.5884323793209852, | |
| "learning_rate": 0.00018591701029393255, | |
| "loss": 0.2898, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.6234908079533132, | |
| "learning_rate": 0.00018579066946908384, | |
| "loss": 0.3411, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.5283455200292801, | |
| "learning_rate": 0.00018566380778638628, | |
| "loss": 0.2715, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.5257190836713216, | |
| "learning_rate": 0.00018553642601605068, | |
| "loss": 0.242, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6940858344290759, | |
| "learning_rate": 0.00018540852493144545, | |
| "loss": 0.3354, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6657062269261629, | |
| "learning_rate": 0.00018528010530909192, | |
| "loss": 0.3073, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5838419940555095, | |
| "learning_rate": 0.00018515116792865957, | |
| "loss": 0.2293, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.6175240728549981, | |
| "learning_rate": 0.00018502171357296144, | |
| "loss": 0.2266, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.514362276581525, | |
| "learning_rate": 0.00018489174302794905, | |
| "loss": 0.2012, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.5495577560324101, | |
| "learning_rate": 0.000184761257082708, | |
| "loss": 0.1992, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.5803859857188628, | |
| "learning_rate": 0.000184630256529453, | |
| "loss": 0.2063, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.6033165262234346, | |
| "learning_rate": 0.00018449874216352306, | |
| "loss": 0.1708, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.6125263532068304, | |
| "learning_rate": 0.00018436671478337666, | |
| "loss": 0.1831, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.5508763917497557, | |
| "learning_rate": 0.00018423417519058694, | |
| "loss": 0.1907, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 0.6343768467745463, | |
| "learning_rate": 0.0001841011241898369, | |
| "loss": 0.2211, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.5592715854666213, | |
| "learning_rate": 0.0001839675625889143, | |
| "loss": 0.1835, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.6853195392487573, | |
| "learning_rate": 0.00018383349119870695, | |
| "loss": 0.2482, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.7078404397255201, | |
| "learning_rate": 0.00018369891083319778, | |
| "loss": 0.2346, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.5364140790780816, | |
| "learning_rate": 0.00018356382230945976, | |
| "loss": 0.2005, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.5699757158649345, | |
| "learning_rate": 0.00018342822644765104, | |
| "loss": 0.1952, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.533449428098461, | |
| "learning_rate": 0.00018329212407100994, | |
| "loss": 0.1827, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.5434329673812834, | |
| "learning_rate": 0.00018315551600585009, | |
| "loss": 0.1977, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.6086249429030556, | |
| "learning_rate": 0.00018301840308155507, | |
| "loss": 0.2075, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.5193552341089737, | |
| "learning_rate": 0.0001828807861305738, | |
| "loss": 0.1774, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 0.5899649673976222, | |
| "learning_rate": 0.00018274266598841517, | |
| "loss": 0.1935, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.6261684110299474, | |
| "learning_rate": 0.0001826040434936431, | |
| "loss": 0.1879, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.5177380809892513, | |
| "learning_rate": 0.0001824649194878714, | |
| "loss": 0.1605, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.6856627216594803, | |
| "learning_rate": 0.00018232529481575872, | |
| "loss": 0.1959, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.7111645847311474, | |
| "learning_rate": 0.00018218517032500344, | |
| "loss": 0.2339, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.6073029325620858, | |
| "learning_rate": 0.00018204454686633834, | |
| "loss": 0.1954, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.5935505620907925, | |
| "learning_rate": 0.00018190342529352565, | |
| "loss": 0.1935, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.6542152124825684, | |
| "learning_rate": 0.0001817618064633518, | |
| "loss": 0.2287, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.5147785285870337, | |
| "learning_rate": 0.0001816196912356222, | |
| "loss": 0.182, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.5727570935780749, | |
| "learning_rate": 0.00018147708047315587, | |
| "loss": 0.1815, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.5657627178730809, | |
| "learning_rate": 0.00018133397504178057, | |
| "loss": 0.2132, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 0.5662562121438198, | |
| "learning_rate": 0.00018119037581032724, | |
| "loss": 0.2056, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.6099744881610398, | |
| "learning_rate": 0.00018104628365062477, | |
| "loss": 0.213, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.5460632540608352, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.162, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.5391352126605721, | |
| "learning_rate": 0.00018075662404874626, | |
| "loss": 0.1893, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.6494284594879935, | |
| "learning_rate": 0.00018061105836517024, | |
| "loss": 0.2291, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.5004570177715079, | |
| "learning_rate": 0.00018046500327053463, | |
| "loss": 0.1541, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.5918866837652731, | |
| "learning_rate": 0.0001803184596515784, | |
| "loss": 0.2188, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.4814928700454864, | |
| "learning_rate": 0.00018017142839800668, | |
| "loss": 0.1371, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_blimp_filtered_avg": 0.7256716417910448, | |
| "eval_blimp_filtered_std": 0.004928318952407523, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_blimp_supplement_avg": 0.8405172413793104, | |
| "eval_blimp_supplement_std": 0.016405077514349695, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_vqa_filtered_avg": 0.52, | |
| "eval_vqa_filtered_std": 0.05021167315686779, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_winoground_filtered_avg": 0.65, | |
| "eval_winoground_filtered_std": 0.047937248544110196, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.5140990035555452, | |
| "learning_rate": 0.0001800239104024851, | |
| "loss": 0.1533, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.4933080278392121, | |
| "learning_rate": 0.0001798759065606345, | |
| "loss": 0.1556, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.5882761858054979, | |
| "learning_rate": 0.00017972741777102523, | |
| "loss": 0.2151, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.5476770873594954, | |
| "learning_rate": 0.00017957844493517213, | |
| "loss": 0.1809, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.6421832624123264, | |
| "learning_rate": 0.0001794289889575286, | |
| "loss": 0.2046, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.5179683138411204, | |
| "learning_rate": 0.0001792790507454815, | |
| "loss": 0.176, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.604477016740788, | |
| "learning_rate": 0.00017912863120934534, | |
| "loss": 0.2174, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.5803547980215563, | |
| "learning_rate": 0.00017897773126235688, | |
| "loss": 0.1599, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.5391766451927411, | |
| "learning_rate": 0.0001788263518206697, | |
| "loss": 0.1867, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.5008335310625588, | |
| "learning_rate": 0.00017867449380334834, | |
| "loss": 0.1543, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.5344073359347221, | |
| "learning_rate": 0.00017852215813236305, | |
| "loss": 0.1836, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.5560594943415273, | |
| "learning_rate": 0.000178369345732584, | |
| "loss": 0.1878, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.4967512656076653, | |
| "learning_rate": 0.00017821605753177562, | |
| "loss": 0.1643, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.4616678862036335, | |
| "learning_rate": 0.00017806229446059124, | |
| "loss": 0.1405, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5366458032488048, | |
| "learning_rate": 0.00017790805745256704, | |
| "loss": 0.1563, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5850965548297408, | |
| "learning_rate": 0.00017775334744411678, | |
| "loss": 0.1735, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.6019045486967092, | |
| "learning_rate": 0.00017759816537452574, | |
| "loss": 0.1702, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.5447191325941722, | |
| "learning_rate": 0.00017744251218594542, | |
| "loss": 0.1536, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.5782504760607824, | |
| "learning_rate": 0.00017728638882338746, | |
| "loss": 0.1765, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.6057705101997692, | |
| "learning_rate": 0.00017712979623471807, | |
| "loss": 0.1771, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.6097931752718987, | |
| "learning_rate": 0.00017697273537065232, | |
| "loss": 0.2139, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.56308892592465, | |
| "learning_rate": 0.00017681520718474823, | |
| "loss": 0.1959, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.48304728057105417, | |
| "learning_rate": 0.00017665721263340113, | |
| "loss": 0.1541, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.5046672087816175, | |
| "learning_rate": 0.0001764987526758377, | |
| "loss": 0.2057, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.591812842303311, | |
| "learning_rate": 0.00017633982827411032, | |
| "loss": 0.2117, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.5124492509013987, | |
| "learning_rate": 0.00017618044039309098, | |
| "loss": 0.1782, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.49461512298625476, | |
| "learning_rate": 0.0001760205900004657, | |
| "loss": 0.1587, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.5072005711040857, | |
| "learning_rate": 0.00017586027806672857, | |
| "loss": 0.1551, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.7166883343661715, | |
| "learning_rate": 0.00017569950556517566, | |
| "loss": 0.2556, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.6256555159933375, | |
| "learning_rate": 0.00017553827347189938, | |
| "loss": 0.183, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.533828362038465, | |
| "learning_rate": 0.00017537658276578247, | |
| "loss": 0.1606, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.5661675325630594, | |
| "learning_rate": 0.00017521443442849188, | |
| "loss": 0.1789, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.5680894670630955, | |
| "learning_rate": 0.00017505182944447316, | |
| "loss": 0.1879, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.4817701078567901, | |
| "learning_rate": 0.00017488876880094413, | |
| "loss": 0.1521, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 0.4821470961031985, | |
| "learning_rate": 0.0001747252534878891, | |
| "loss": 0.1463, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.49328987347131154, | |
| "learning_rate": 0.0001745612844980528, | |
| "loss": 0.1717, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5124773897350241, | |
| "learning_rate": 0.00017439686282693436, | |
| "loss": 0.1664, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.4947058789125862, | |
| "learning_rate": 0.00017423198947278117, | |
| "loss": 0.155, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.49669905069620973, | |
| "learning_rate": 0.00017406666543658304, | |
| "loss": 0.1317, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.5945246572487285, | |
| "learning_rate": 0.00017390089172206592, | |
| "loss": 0.179, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.6088690966027199, | |
| "learning_rate": 0.00017373466933568588, | |
| "loss": 0.1688, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.5504854313763279, | |
| "learning_rate": 0.00017356799928662297, | |
| "loss": 0.1409, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.5616999197359607, | |
| "learning_rate": 0.00017340088258677522, | |
| "loss": 0.1611, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.5630261456683686, | |
| "learning_rate": 0.00017323332025075223, | |
| "loss": 0.1424, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.5286011853132182, | |
| "learning_rate": 0.00017306531329586933, | |
| "loss": 0.1435, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.5614256677889038, | |
| "learning_rate": 0.00017289686274214118, | |
| "loss": 0.1717, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.5877414440813399, | |
| "learning_rate": 0.00017272796961227563, | |
| "loss": 0.15, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.5368624648164415, | |
| "learning_rate": 0.00017255863493166756, | |
| "loss": 0.1549, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.5500707727315131, | |
| "learning_rate": 0.0001723888597283926, | |
| "loss": 0.175, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.6833139983554208, | |
| "learning_rate": 0.00017221864503320092, | |
| "loss": 0.1678, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.5282683935361436, | |
| "learning_rate": 0.00017204799187951105, | |
| "loss": 0.1464, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.6220938804700495, | |
| "learning_rate": 0.00017187690130340328, | |
| "loss": 0.1936, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.5368424100980813, | |
| "learning_rate": 0.00017170537434361386, | |
| "loss": 0.16, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.4964367798749474, | |
| "learning_rate": 0.0001715334120415283, | |
| "loss": 0.1315, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.5332125576256961, | |
| "learning_rate": 0.00017136101544117525, | |
| "loss": 0.1601, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.6452241501826331, | |
| "learning_rate": 0.00017118818558922003, | |
| "loss": 0.1944, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.5478001238846428, | |
| "learning_rate": 0.00017101492353495845, | |
| "loss": 0.1628, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.5907872420921463, | |
| "learning_rate": 0.00017084123033031024, | |
| "loss": 0.1979, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.5539786918398002, | |
| "learning_rate": 0.0001706671070298128, | |
| "loss": 0.1569, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.504557687915929, | |
| "learning_rate": 0.00017049255469061474, | |
| "loss": 0.1676, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.5193398413523059, | |
| "learning_rate": 0.00017031757437246947, | |
| "loss": 0.1535, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.5089502238747983, | |
| "learning_rate": 0.00017014216713772884, | |
| "loss": 0.1609, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.5090619290397491, | |
| "learning_rate": 0.00016996633405133655, | |
| "loss": 0.1601, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.5294032162956256, | |
| "learning_rate": 0.00016979007618082175, | |
| "loss": 0.1416, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.5376905990718213, | |
| "learning_rate": 0.0001696133945962927, | |
| "loss": 0.1704, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.6016988384720686, | |
| "learning_rate": 0.0001694362903704299, | |
| "loss": 0.1724, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.6471404474003535, | |
| "learning_rate": 0.00016925876457848, | |
| "loss": 0.1893, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.5121222508453968, | |
| "learning_rate": 0.00016908081829824912, | |
| "loss": 0.1166, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.5997272693042293, | |
| "learning_rate": 0.0001689024526100961, | |
| "loss": 0.2107, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.5636921436981437, | |
| "learning_rate": 0.00016872366859692627, | |
| "loss": 0.1805, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.6126951187530087, | |
| "learning_rate": 0.00016854446734418466, | |
| "loss": 0.1914, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.554597312718397, | |
| "learning_rate": 0.0001683648499398495, | |
| "loss": 0.1836, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.5003434649301144, | |
| "learning_rate": 0.00016818481747442554, | |
| "loss": 0.1631, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.5181592112650633, | |
| "learning_rate": 0.0001680043710409375, | |
| "loss": 0.1323, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.5773215207200062, | |
| "learning_rate": 0.00016782351173492342, | |
| "loss": 0.2156, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.46149339744192175, | |
| "learning_rate": 0.00016764224065442796, | |
| "loss": 0.1308, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.5154009122145548, | |
| "learning_rate": 0.0001674605588999959, | |
| "loss": 0.1723, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.5032409831088436, | |
| "learning_rate": 0.0001672784675746651, | |
| "loss": 0.1522, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.5765338444046834, | |
| "learning_rate": 0.00016709596778396026, | |
| "loss": 0.1746, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.6731025984083968, | |
| "learning_rate": 0.00016691306063588583, | |
| "loss": 0.2205, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.4837466365127126, | |
| "learning_rate": 0.00016672974724091954, | |
| "loss": 0.1553, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.5424268673951455, | |
| "learning_rate": 0.00016654602871200546, | |
| "loss": 0.1433, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.5171465106257717, | |
| "learning_rate": 0.0001663619061645474, | |
| "loss": 0.1698, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.5543423821478913, | |
| "learning_rate": 0.00016617738071640208, | |
| "loss": 0.1494, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.5614170511875883, | |
| "learning_rate": 0.0001659924534878723, | |
| "loss": 0.1719, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.5817955693629092, | |
| "learning_rate": 0.0001658071256017001, | |
| "loss": 0.1653, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.5093939905856546, | |
| "learning_rate": 0.0001656213981830602, | |
| "loss": 0.1331, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.530369374719384, | |
| "learning_rate": 0.00016543527235955282, | |
| "loss": 0.1703, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.5540917573532339, | |
| "learning_rate": 0.00016524874926119717, | |
| "loss": 0.1801, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.5189763553195834, | |
| "learning_rate": 0.0001650618300204242, | |
| "loss": 0.1526, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.4919336835652238, | |
| "learning_rate": 0.00016487451577207018, | |
| "loss": 0.1251, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.5303366245761384, | |
| "learning_rate": 0.00016468680765336936, | |
| "loss": 0.1504, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.579006149147232, | |
| "learning_rate": 0.00016449870680394747, | |
| "loss": 0.2013, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.5698441995370652, | |
| "learning_rate": 0.0001643102143658145, | |
| "loss": 0.1494, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.45335914900667046, | |
| "learning_rate": 0.00016412133148335784, | |
| "loss": 0.1196, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.500799551817792, | |
| "learning_rate": 0.0001639320593033355, | |
| "loss": 0.1576, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.6003518072235265, | |
| "learning_rate": 0.000163742398974869, | |
| "loss": 0.1584, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 0.6168365552101142, | |
| "learning_rate": 0.00016355235164943626, | |
| "loss": 0.2091, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.5713474186129687, | |
| "learning_rate": 0.0001633619184808649, | |
| "loss": 0.1567, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.5874214731593885, | |
| "learning_rate": 0.0001631711006253251, | |
| "loss": 0.1574, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_blimp_filtered_avg": 0.7201492537313433, | |
| "eval_blimp_filtered_std": 0.004978357158642791, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_blimp_supplement_avg": 0.8297413793103449, | |
| "eval_blimp_supplement_std": 0.01657166693464671, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_vqa_filtered_avg": 0.49, | |
| "eval_vqa_filtered_std": 0.05024183937956912, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_winoground_filtered_avg": 0.61, | |
| "eval_winoground_filtered_std": 0.04902071300001975, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.6042802091257128, | |
| "learning_rate": 0.00016297989924132252, | |
| "loss": 0.1818, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.47364098207681, | |
| "learning_rate": 0.00016278831548969134, | |
| "loss": 0.1328, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.5471235998749361, | |
| "learning_rate": 0.00016259635053358717, | |
| "loss": 0.1507, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.5880254357081499, | |
| "learning_rate": 0.00016240400553848007, | |
| "loss": 0.1797, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.4946694660134096, | |
| "learning_rate": 0.0001622112816721474, | |
| "loss": 0.1445, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.5391138081584693, | |
| "learning_rate": 0.0001620181801046667, | |
| "loss": 0.1592, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.4600495974038577, | |
| "learning_rate": 0.00016182470200840868, | |
| "loss": 0.1255, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.4600858105564683, | |
| "learning_rate": 0.00016163084855803006, | |
| "loss": 0.1274, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.4869984031862606, | |
| "learning_rate": 0.00016143662093046638, | |
| "loss": 0.1312, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.5760667761271687, | |
| "learning_rate": 0.000161242020304925, | |
| "loss": 0.1956, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.4914141963984456, | |
| "learning_rate": 0.0001610470478628778, | |
| "loss": 0.1305, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.5844115229455148, | |
| "learning_rate": 0.00016085170478805395, | |
| "loss": 0.1767, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.5609976894365024, | |
| "learning_rate": 0.00016065599226643303, | |
| "loss": 0.1568, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.5199227181198156, | |
| "learning_rate": 0.0001604599114862375, | |
| "loss": 0.1358, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.580620607781089, | |
| "learning_rate": 0.00016026346363792567, | |
| "loss": 0.1568, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.5406448317458065, | |
| "learning_rate": 0.00016006664991418434, | |
| "loss": 0.1398, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.44523867426706976, | |
| "learning_rate": 0.00015986947150992172, | |
| "loss": 0.1355, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.5115419137782133, | |
| "learning_rate": 0.0001596719296222601, | |
| "loss": 0.1468, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.5985959567307599, | |
| "learning_rate": 0.0001594740254505285, | |
| "loss": 0.1378, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.6222671216404903, | |
| "learning_rate": 0.0001592757601962555, | |
| "loss": 0.184, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5413250088547485, | |
| "learning_rate": 0.00015907713506316192, | |
| "loss": 0.1758, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5382013654400807, | |
| "learning_rate": 0.00015887815125715344, | |
| "loss": 0.1509, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.48249026147532587, | |
| "learning_rate": 0.00015867880998631347, | |
| "loss": 0.1183, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.5223760532607825, | |
| "learning_rate": 0.0001584791124608955, | |
| "loss": 0.1446, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.42511278258760743, | |
| "learning_rate": 0.0001582790598933161, | |
| "loss": 0.1085, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.5280124307363728, | |
| "learning_rate": 0.00015807865349814733, | |
| "loss": 0.1211, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.44071272749150264, | |
| "learning_rate": 0.00015787789449210938, | |
| "loss": 0.1199, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.5672982836442205, | |
| "learning_rate": 0.0001576767840940633, | |
| "loss": 0.1653, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.5407137471822783, | |
| "learning_rate": 0.00015747532352500357, | |
| "loss": 0.1459, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.47958987114022433, | |
| "learning_rate": 0.00015727351400805052, | |
| "loss": 0.122, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.4949008797838833, | |
| "learning_rate": 0.0001570713567684432, | |
| "loss": 0.1476, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 0.5706163209211994, | |
| "learning_rate": 0.0001568688530335316, | |
| "loss": 0.1859, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.48156623541863475, | |
| "learning_rate": 0.0001566660040327695, | |
| "loss": 0.1354, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.48592626955333434, | |
| "learning_rate": 0.00015646281099770682, | |
| "loss": 0.1287, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.42812001499236796, | |
| "learning_rate": 0.00015625927516198232, | |
| "loss": 0.1041, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.5351459013098889, | |
| "learning_rate": 0.0001560553977613158, | |
| "loss": 0.1321, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.603892324255596, | |
| "learning_rate": 0.00015585118003350092, | |
| "loss": 0.1524, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.5704162659549111, | |
| "learning_rate": 0.00015564662321839755, | |
| "loss": 0.1643, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.5619464269808508, | |
| "learning_rate": 0.00015544172855792423, | |
| "loss": 0.1695, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.5552857752448269, | |
| "learning_rate": 0.0001552364972960506, | |
| "loss": 0.1436, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.4859084061375966, | |
| "learning_rate": 0.00015503093067878996, | |
| "loss": 0.1304, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.44977231534740475, | |
| "learning_rate": 0.00015482502995419167, | |
| "loss": 0.1328, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.4713213980355454, | |
| "learning_rate": 0.0001546187963723334, | |
| "loss": 0.1105, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.5341277409726155, | |
| "learning_rate": 0.00015441223118531388, | |
| "loss": 0.1449, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 0.5705390112616555, | |
| "learning_rate": 0.00015420533564724495, | |
| "loss": 0.1695, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.4774790901030609, | |
| "learning_rate": 0.00015399811101424418, | |
| "loss": 0.1403, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.5512200465464098, | |
| "learning_rate": 0.00015379055854442708, | |
| "loss": 0.1643, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.5219733540585905, | |
| "learning_rate": 0.00015358267949789966, | |
| "loss": 0.1237, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.5033089815712022, | |
| "learning_rate": 0.0001533744751367506, | |
| "loss": 0.1291, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.4747834365201173, | |
| "learning_rate": 0.0001531659467250436, | |
| "loss": 0.1275, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 0.40965665913605903, | |
| "learning_rate": 0.0001529570955288099, | |
| "loss": 0.1156, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.440005816758666, | |
| "learning_rate": 0.00015274792281604028, | |
| "loss": 0.1099, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.5579574679556536, | |
| "learning_rate": 0.00015253842985667762, | |
| "loss": 0.1456, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.4959791770798946, | |
| "learning_rate": 0.0001523286179226091, | |
| "loss": 0.1468, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.5066322640800116, | |
| "learning_rate": 0.0001521184882876585, | |
| "loss": 0.1085, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.5297332606909233, | |
| "learning_rate": 0.00015190804222757845, | |
| "loss": 0.1476, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.5830381345482916, | |
| "learning_rate": 0.00015169728102004256, | |
| "loss": 0.1424, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.44997991084618094, | |
| "learning_rate": 0.00015148620594463794, | |
| "loss": 0.1068, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.5459028855766525, | |
| "learning_rate": 0.00015127481828285718, | |
| "loss": 0.1673, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.46924015811236747, | |
| "learning_rate": 0.0001510631193180907, | |
| "loss": 0.1136, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.5011418751892371, | |
| "learning_rate": 0.00015085111033561895, | |
| "loss": 0.1154, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.5328628999011582, | |
| "learning_rate": 0.00015063879262260446, | |
| "loss": 0.1397, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.4964913122706592, | |
| "learning_rate": 0.00015042616746808435, | |
| "loss": 0.144, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.501470133961365, | |
| "learning_rate": 0.00015021323616296213, | |
| "loss": 0.1379, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.5269212199791002, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.1311, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.5049669772855443, | |
| "learning_rate": 0.00014978646027381123, | |
| "loss": 0.1196, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.4525225552090921, | |
| "learning_rate": 0.00014957261828085191, | |
| "loss": 0.1141, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.4790059846117726, | |
| "learning_rate": 0.0001493584753194134, | |
| "loss": 0.1028, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.5832150044411881, | |
| "learning_rate": 0.00014914403268961426, | |
| "loss": 0.158, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.522112830999454, | |
| "learning_rate": 0.00014892929169339235, | |
| "loss": 0.1459, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.4887536825683872, | |
| "learning_rate": 0.00014871425363449718, | |
| "loss": 0.1263, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.467502684654505, | |
| "learning_rate": 0.00014849891981848158, | |
| "loss": 0.1388, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.4546537818819024, | |
| "learning_rate": 0.0001482832915526942, | |
| "loss": 0.105, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.541602974136478, | |
| "learning_rate": 0.00014806737014627124, | |
| "loss": 0.1388, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.5066261091776695, | |
| "learning_rate": 0.00014785115691012864, | |
| "loss": 0.1231, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.6064880991698398, | |
| "learning_rate": 0.00014763465315695425, | |
| "loss": 0.173, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.5259754097567955, | |
| "learning_rate": 0.00014741786020119955, | |
| "loss": 0.1383, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.5335061292186741, | |
| "learning_rate": 0.00014720077935907196, | |
| "loss": 0.1671, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 0.43421172131647107, | |
| "learning_rate": 0.00014698341194852664, | |
| "loss": 0.1207, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.4476149707237085, | |
| "learning_rate": 0.00014676575928925867, | |
| "loss": 0.1103, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.49116627439944716, | |
| "learning_rate": 0.00014654782270269497, | |
| "loss": 0.1195, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.44289326248169675, | |
| "learning_rate": 0.00014632960351198618, | |
| "loss": 0.1217, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.4022818703819969, | |
| "learning_rate": 0.00014611110304199872, | |
| "loss": 0.0868, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.43714254165230704, | |
| "learning_rate": 0.00014589232261930674, | |
| "loss": 0.1062, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 0.4499910364889525, | |
| "learning_rate": 0.00014567326357218407, | |
| "loss": 0.1115, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.4812264446699121, | |
| "learning_rate": 0.00014545392723059616, | |
| "loss": 0.1272, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.5466847826869639, | |
| "learning_rate": 0.0001452343149261919, | |
| "loss": 0.1387, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.4671789379009673, | |
| "learning_rate": 0.00014501442799229572, | |
| "loss": 0.1227, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.4168568601736311, | |
| "learning_rate": 0.00014479426776389936, | |
| "loss": 0.0955, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.4236347478067532, | |
| "learning_rate": 0.00014457383557765386, | |
| "loss": 0.1226, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.49129515437526694, | |
| "learning_rate": 0.00014435313277186125, | |
| "loss": 0.1411, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.4395945017572998, | |
| "learning_rate": 0.00014413216068646668, | |
| "loss": 0.1306, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 0.4832622577980266, | |
| "learning_rate": 0.0001439109206630501, | |
| "loss": 0.1133, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.4840123051282916, | |
| "learning_rate": 0.0001436894140448183, | |
| "loss": 0.1354, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.6106816446433545, | |
| "learning_rate": 0.00014346764217659653, | |
| "loss": 0.1759, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.4439633620623163, | |
| "learning_rate": 0.0001432456064048204, | |
| "loss": 0.105, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.6103065192992649, | |
| "learning_rate": 0.00014302330807752786, | |
| "loss": 0.1736, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.4594414148489128, | |
| "learning_rate": 0.0001428007485443509, | |
| "loss": 0.1067, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.48309766302252394, | |
| "learning_rate": 0.00014257792915650728, | |
| "loss": 0.1319, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.4979398119598126, | |
| "learning_rate": 0.00014235485126679243, | |
| "loss": 0.1361, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_blimp_filtered_avg": 0.7231343283582089, | |
| "eval_blimp_filtered_std": 0.004963011033511667, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_blimp_supplement_avg": 0.8275862068965517, | |
| "eval_blimp_supplement_std": 0.0167403676680407, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_vqa_filtered_avg": 0.47, | |
| "eval_vqa_filtered_std": 0.0501613558046592, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_winoground_filtered_avg": 0.64, | |
| "eval_winoground_filtered_std": 0.048241815132442176, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.4877241025128464, | |
| "learning_rate": 0.00014213151622957128, | |
| "loss": 0.1502, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.5812751924824102, | |
| "learning_rate": 0.00014190792540076986, | |
| "loss": 0.1687, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 0.4627935359645996, | |
| "learning_rate": 0.00014168408013786728, | |
| "loss": 0.1246, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4587837001229827, | |
| "learning_rate": 0.00014145998179988735, | |
| "loss": 0.1072, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5196182852163257, | |
| "learning_rate": 0.00014123563174739037, | |
| "loss": 0.1269, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.42954431425652795, | |
| "learning_rate": 0.0001410110313424648, | |
| "loss": 0.1034, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.4551597635457604, | |
| "learning_rate": 0.00014078618194871914, | |
| "loss": 0.1084, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 0.5751209424815239, | |
| "learning_rate": 0.0001405610849312736, | |
| "loss": 0.136, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.5118033581542103, | |
| "learning_rate": 0.00014033574165675164, | |
| "loss": 0.1241, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.43644759709330094, | |
| "learning_rate": 0.00014011015349327187, | |
| "loss": 0.117, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.4958861591526709, | |
| "learning_rate": 0.00013988432181043982, | |
| "loss": 0.1163, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.44950223725406846, | |
| "learning_rate": 0.00013965824797933926, | |
| "loss": 0.1013, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.5174784327558367, | |
| "learning_rate": 0.0001394319333725243, | |
| "loss": 0.1228, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.451176763896539, | |
| "learning_rate": 0.00013920537936401077, | |
| "loss": 0.1137, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.35979415227278505, | |
| "learning_rate": 0.00013897858732926793, | |
| "loss": 0.0893, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.4087154672782133, | |
| "learning_rate": 0.0001387515586452103, | |
| "loss": 0.0979, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.4238917271153031, | |
| "learning_rate": 0.000138524294690189, | |
| "loss": 0.0902, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.37289193742062965, | |
| "learning_rate": 0.00013829679684398375, | |
| "loss": 0.0905, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.4561111955028976, | |
| "learning_rate": 0.000138069066487794, | |
| "loss": 0.1154, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.4795856880336275, | |
| "learning_rate": 0.00013784110500423104, | |
| "loss": 0.1053, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.5758379418361174, | |
| "learning_rate": 0.00013761291377730936, | |
| "loss": 0.1646, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.3967821645775129, | |
| "learning_rate": 0.00013738449419243827, | |
| "loss": 0.0797, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.5352641990707583, | |
| "learning_rate": 0.00013715584763641345, | |
| "loss": 0.1205, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.39673398503322155, | |
| "learning_rate": 0.0001369269754974087, | |
| "loss": 0.0672, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5673600812406742, | |
| "learning_rate": 0.00013669787916496722, | |
| "loss": 0.1269, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5134798249291933, | |
| "learning_rate": 0.00013646856002999354, | |
| "loss": 0.1275, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.26137163241501854, | |
| "learning_rate": 0.00013623901948474473, | |
| "loss": 0.0391, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.3318640399705981, | |
| "learning_rate": 0.00013600925892282218, | |
| "loss": 0.0414, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.39323951881930796, | |
| "learning_rate": 0.00013577927973916306, | |
| "loss": 0.054, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.31507008260792096, | |
| "learning_rate": 0.0001355490833300318, | |
| "loss": 0.042, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.351605629418437, | |
| "learning_rate": 0.00013531867109301175, | |
| "loss": 0.0419, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.300514728252508, | |
| "learning_rate": 0.00013508804442699648, | |
| "loss": 0.0432, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.429759783149188, | |
| "learning_rate": 0.00013485720473218154, | |
| "loss": 0.0434, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.3371850480060168, | |
| "learning_rate": 0.00013462615341005573, | |
| "loss": 0.0398, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.41727641506204793, | |
| "learning_rate": 0.00013439489186339282, | |
| "loss": 0.0532, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.4526462472237055, | |
| "learning_rate": 0.0001341634214962428, | |
| "loss": 0.0504, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.394759014800421, | |
| "learning_rate": 0.00013393174371392348, | |
| "loss": 0.0362, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.43530217979086694, | |
| "learning_rate": 0.00013369985992301198, | |
| "loss": 0.0581, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.3489279774411472, | |
| "learning_rate": 0.00013346777153133615, | |
| "loss": 0.0428, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.3524007670573337, | |
| "learning_rate": 0.00013323547994796597, | |
| "loss": 0.0392, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.4161180575138653, | |
| "learning_rate": 0.00013300298658320517, | |
| "loss": 0.0435, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.36047484402484764, | |
| "learning_rate": 0.00013277029284858237, | |
| "loss": 0.0453, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.3660355277365804, | |
| "learning_rate": 0.00013253740015684284, | |
| "loss": 0.0442, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.3239828127697563, | |
| "learning_rate": 0.00013230430992193973, | |
| "loss": 0.0405, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 0.4158448690843572, | |
| "learning_rate": 0.00013207102355902552, | |
| "loss": 0.0524, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.27764986772544564, | |
| "learning_rate": 0.00013183754248444343, | |
| "loss": 0.0332, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.32771023150803125, | |
| "learning_rate": 0.00013160386811571876, | |
| "loss": 0.0459, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.29940107859442794, | |
| "learning_rate": 0.0001313700018715505, | |
| "loss": 0.0377, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 0.3819582084310897, | |
| "learning_rate": 0.00013113594517180242, | |
| "loss": 0.058, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 0.32714441974400016, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.0437, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.36220473718120416, | |
| "learning_rate": 0.00013066726609079526, | |
| "loss": 0.0425, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.3426348230655199, | |
| "learning_rate": 0.00013043264655501074, | |
| "loss": 0.0396, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.32235085289402154, | |
| "learning_rate": 0.00013019784225457855, | |
| "loss": 0.0382, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.2964086383376768, | |
| "learning_rate": 0.0001299628546150577, | |
| "loss": 0.0382, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.27015454300882835, | |
| "learning_rate": 0.00012972768506312027, | |
| "loss": 0.0283, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 0.33035662649730047, | |
| "learning_rate": 0.00012949233502654284, | |
| "loss": 0.0384, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.3187428949344172, | |
| "learning_rate": 0.00012925680593419778, | |
| "loss": 0.032, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.3399662779151198, | |
| "learning_rate": 0.00012902109921604448, | |
| "loss": 0.0405, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.35365859482095496, | |
| "learning_rate": 0.00012878521630312078, | |
| "loss": 0.0397, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.30675985331556893, | |
| "learning_rate": 0.00012854915862753422, | |
| "loss": 0.035, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.414074353684469, | |
| "learning_rate": 0.0001283129276224534, | |
| "loss": 0.052, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.38344430660186296, | |
| "learning_rate": 0.0001280765247220993, | |
| "loss": 0.0359, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.3137291177532792, | |
| "learning_rate": 0.0001278399513617364, | |
| "loss": 0.031, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.39542817136382147, | |
| "learning_rate": 0.0001276032089776642, | |
| "loss": 0.0438, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.3778319793304532, | |
| "learning_rate": 0.0001273662990072083, | |
| "loss": 0.0485, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.35242405330198695, | |
| "learning_rate": 0.0001271292228887118, | |
| "loss": 0.0395, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.24773421468492, | |
| "learning_rate": 0.00012689198206152657, | |
| "loss": 0.0205, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.37351452341587044, | |
| "learning_rate": 0.00012665457796600443, | |
| "loss": 0.0403, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.29584266385998476, | |
| "learning_rate": 0.0001264170120434884, | |
| "loss": 0.0332, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.28547841885732445, | |
| "learning_rate": 0.00012617928573630406, | |
| "loss": 0.0288, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.37964584553416536, | |
| "learning_rate": 0.0001259414004877507, | |
| "loss": 0.0348, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 0.36767195929723195, | |
| "learning_rate": 0.0001257033577420926, | |
| "loss": 0.0404, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.3509417575248461, | |
| "learning_rate": 0.00012546515894455026, | |
| "loss": 0.0373, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.2848574183498485, | |
| "learning_rate": 0.00012522680554129156, | |
| "loss": 0.0338, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.3019415495988161, | |
| "learning_rate": 0.0001249882989794231, | |
| "loss": 0.0342, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.2990668841044976, | |
| "learning_rate": 0.00012474964070698127, | |
| "loss": 0.0335, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.4328132625325146, | |
| "learning_rate": 0.00012451083217292357, | |
| "loss": 0.0394, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.4292821742979654, | |
| "learning_rate": 0.00012427187482711986, | |
| "loss": 0.0516, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.2686010674908365, | |
| "learning_rate": 0.0001240327701203433, | |
| "loss": 0.0344, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.28452697548238837, | |
| "learning_rate": 0.00012379351950426187, | |
| "loss": 0.0307, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.3597827634214583, | |
| "learning_rate": 0.00012355412443142936, | |
| "loss": 0.0429, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.31337666682629667, | |
| "learning_rate": 0.00012331458635527658, | |
| "loss": 0.038, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.3751035825027268, | |
| "learning_rate": 0.0001230749067301025, | |
| "loss": 0.0385, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.3982139708950298, | |
| "learning_rate": 0.00012283508701106557, | |
| "loss": 0.0461, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.3672383693603079, | |
| "learning_rate": 0.00012259512865417477, | |
| "loss": 0.045, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.3836978621116552, | |
| "learning_rate": 0.00012235503311628073, | |
| "loss": 0.0434, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.3661127438866315, | |
| "learning_rate": 0.00012211480185506698, | |
| "loss": 0.0374, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.39751744692779245, | |
| "learning_rate": 0.00012187443632904105, | |
| "loss": 0.0334, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.2943198017134509, | |
| "learning_rate": 0.00012163393799752565, | |
| "loss": 0.0354, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.3002324681197963, | |
| "learning_rate": 0.00012139330832064974, | |
| "loss": 0.0263, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.29049707177485556, | |
| "learning_rate": 0.00012115254875933979, | |
| "loss": 0.0374, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.36487904735548193, | |
| "learning_rate": 0.00012091166077531075, | |
| "loss": 0.0366, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.6001906832098756, | |
| "learning_rate": 0.00012067064583105729, | |
| "loss": 0.0397, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.29594265959931015, | |
| "learning_rate": 0.00012042950538984492, | |
| "loss": 0.0339, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.3136944220749086, | |
| "learning_rate": 0.00012018824091570103, | |
| "loss": 0.0383, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.34881594693482015, | |
| "learning_rate": 0.00011994685387340607, | |
| "loss": 0.0328, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.343330706863882, | |
| "learning_rate": 0.00011970534572848464, | |
| "loss": 0.0402, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.34079159501332024, | |
| "learning_rate": 0.00011946371794719656, | |
| "loss": 0.0351, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 0.2827419424824256, | |
| "learning_rate": 0.000119221971996528, | |
| "loss": 0.0257, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 0.3123401643495521, | |
| "learning_rate": 0.0001189801093441826, | |
| "loss": 0.029, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_blimp_filtered_avg": 0.7173134328358209, | |
| "eval_blimp_filtered_std": 0.005026688908914533, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_blimp_supplement_avg": 0.8254310344827587, | |
| "eval_blimp_supplement_std": 0.017008878963692253, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_vqa_filtered_avg": 0.5, | |
| "eval_vqa_filtered_std": 0.050251890762960605, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_winoground_filtered_avg": 0.65, | |
| "eval_winoground_filtered_std": 0.04793724854411019, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.3284360674259941, | |
| "learning_rate": 0.00011873813145857249, | |
| "loss": 0.0307, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.4088320006622324, | |
| "learning_rate": 0.0001184960398088094, | |
| "loss": 0.0424, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.3347370845203778, | |
| "learning_rate": 0.00011825383586469583, | |
| "loss": 0.0301, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.34998604945319767, | |
| "learning_rate": 0.00011801152109671595, | |
| "loss": 0.0398, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.28072401871679603, | |
| "learning_rate": 0.00011776909697602689, | |
| "loss": 0.0207, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.27417539486849196, | |
| "learning_rate": 0.00011752656497444952, | |
| "loss": 0.0303, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 0.3528956989148332, | |
| "learning_rate": 0.00011728392656445981, | |
| "loss": 0.035, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 0.2732570327379083, | |
| "learning_rate": 0.00011704118321917976, | |
| "loss": 0.0293, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.23173712745705208, | |
| "learning_rate": 0.00011679833641236844, | |
| "loss": 0.0194, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.3833499067600359, | |
| "learning_rate": 0.000116555387618413, | |
| "loss": 0.0301, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.3445447080613637, | |
| "learning_rate": 0.00011631233831231991, | |
| "loss": 0.0394, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.3521386004172123, | |
| "learning_rate": 0.00011606918996970573, | |
| "loss": 0.0359, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.32961191495589676, | |
| "learning_rate": 0.00011582594406678839, | |
| "loss": 0.0353, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.32438381236368063, | |
| "learning_rate": 0.00011558260208037817, | |
| "loss": 0.0326, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.31086642828231076, | |
| "learning_rate": 0.00011533916548786857, | |
| "loss": 0.0319, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.3725691312752257, | |
| "learning_rate": 0.00011509563576722753, | |
| "loss": 0.0419, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.2712234441709734, | |
| "learning_rate": 0.00011485201439698847, | |
| "loss": 0.0286, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.2992546512802102, | |
| "learning_rate": 0.00011460830285624118, | |
| "loss": 0.0298, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.3740397615866031, | |
| "learning_rate": 0.00011436450262462284, | |
| "loss": 0.0434, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.34323442855123726, | |
| "learning_rate": 0.00011412061518230914, | |
| "loss": 0.038, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.2859959872205353, | |
| "learning_rate": 0.00011387664201000532, | |
| "loss": 0.0337, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 0.35126159863306816, | |
| "learning_rate": 0.00011363258458893699, | |
| "loss": 0.0418, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 0.36586330790757604, | |
| "learning_rate": 0.00011338844440084138, | |
| "loss": 0.0472, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 0.3515588484900087, | |
| "learning_rate": 0.0001131442229279581, | |
| "loss": 0.0343, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.26331852309035336, | |
| "learning_rate": 0.00011289992165302035, | |
| "loss": 0.0269, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.304600621064075, | |
| "learning_rate": 0.00011265554205924575, | |
| "loss": 0.0312, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.29191144254029494, | |
| "learning_rate": 0.0001124110856303275, | |
| "loss": 0.0351, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 0.39129462924046776, | |
| "learning_rate": 0.00011216655385042525, | |
| "loss": 0.035, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 0.38533876670236805, | |
| "learning_rate": 0.00011192194820415609, | |
| "loss": 0.0398, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.2395517368524462, | |
| "learning_rate": 0.00011167727017658562, | |
| "loss": 0.0305, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.2882756076173429, | |
| "learning_rate": 0.00011143252125321892, | |
| "loss": 0.028, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3143279186770457, | |
| "learning_rate": 0.00011118770291999137, | |
| "loss": 0.0349, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.2851418078470389, | |
| "learning_rate": 0.00011094281666325988, | |
| "loss": 0.0274, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.36040663639869686, | |
| "learning_rate": 0.00011069786396979367, | |
| "loss": 0.0357, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.33153007014041297, | |
| "learning_rate": 0.00011045284632676536, | |
| "loss": 0.0339, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.28195191033726724, | |
| "learning_rate": 0.00011020776522174186, | |
| "loss": 0.0265, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.29390455843884594, | |
| "learning_rate": 0.0001099626221426754, | |
| "loss": 0.031, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.38900881293341505, | |
| "learning_rate": 0.0001097174185778945, | |
| "loss": 0.0277, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.3393893043389277, | |
| "learning_rate": 0.00010947215601609479, | |
| "loss": 0.0373, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.2908885670583566, | |
| "learning_rate": 0.00010922683594633021, | |
| "loss": 0.0275, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.32162535341655224, | |
| "learning_rate": 0.00010898145985800381, | |
| "loss": 0.0296, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.3101819804262916, | |
| "learning_rate": 0.00010873602924085869, | |
| "loss": 0.0279, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.3225846056983095, | |
| "learning_rate": 0.00010849054558496905, | |
| "loss": 0.0359, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.3320709262623266, | |
| "learning_rate": 0.00010824501038073116, | |
| "loss": 0.0272, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.34267032779621814, | |
| "learning_rate": 0.00010799942511885418, | |
| "loss": 0.0349, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.3252434307442351, | |
| "learning_rate": 0.00010775379129035116, | |
| "loss": 0.0353, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.3424342161187399, | |
| "learning_rate": 0.00010750811038653008, | |
| "loss": 0.0395, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.3350229460047768, | |
| "learning_rate": 0.00010726238389898471, | |
| "loss": 0.0311, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.3333212059527241, | |
| "learning_rate": 0.00010701661331958553, | |
| "loss": 0.0305, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.342236697340712, | |
| "learning_rate": 0.00010677080014047076, | |
| "loss": 0.0387, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.31813476077439445, | |
| "learning_rate": 0.00010652494585403725, | |
| "loss": 0.0332, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.34237738807094575, | |
| "learning_rate": 0.00010627905195293135, | |
| "loss": 0.0305, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.40607599891255564, | |
| "learning_rate": 0.00010603311993004004, | |
| "loss": 0.0394, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.29414538244181715, | |
| "learning_rate": 0.00010578715127848167, | |
| "loss": 0.03, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.37329000324002554, | |
| "learning_rate": 0.000105541147491597, | |
| "loss": 0.0421, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 0.3522250044405319, | |
| "learning_rate": 0.00010529511006294009, | |
| "loss": 0.045, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.28121115420821313, | |
| "learning_rate": 0.00010504904048626925, | |
| "loss": 0.0292, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.26325089300460475, | |
| "learning_rate": 0.00010480294025553798, | |
| "loss": 0.0292, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.2719725064408085, | |
| "learning_rate": 0.00010455681086488586, | |
| "loss": 0.0231, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.3453111024050063, | |
| "learning_rate": 0.00010431065380862959, | |
| "loss": 0.0378, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.27775866655685405, | |
| "learning_rate": 0.00010406447058125368, | |
| "loss": 0.0286, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.33965483172047317, | |
| "learning_rate": 0.00010381826267740171, | |
| "loss": 0.0333, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.27735302965276887, | |
| "learning_rate": 0.00010357203159186694, | |
| "loss": 0.0271, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.3359647684514768, | |
| "learning_rate": 0.0001033257788195835, | |
| "loss": 0.0399, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 0.31390916758505816, | |
| "learning_rate": 0.00010307950585561706, | |
| "loss": 0.0259, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 0.26289076188241683, | |
| "learning_rate": 0.0001028332141951559, | |
| "loss": 0.0303, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.4074393150808, | |
| "learning_rate": 0.0001025869053335019, | |
| "loss": 0.0339, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.254687354277676, | |
| "learning_rate": 0.0001023405807660613, | |
| "loss": 0.0263, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.31504128834532047, | |
| "learning_rate": 0.0001020942419883357, | |
| "loss": 0.0308, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 0.3328845285750464, | |
| "learning_rate": 0.00010184789049591299, | |
| "loss": 0.0355, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 0.31546120267640665, | |
| "learning_rate": 0.00010160152778445829, | |
| "loss": 0.0352, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.2983572052164693, | |
| "learning_rate": 0.0001013551553497047, | |
| "loss": 0.0293, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.31011068039080025, | |
| "learning_rate": 0.0001011087746874445, | |
| "loss": 0.0354, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.2987245112804117, | |
| "learning_rate": 0.00010086238729351988, | |
| "loss": 0.0354, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.3234038283715016, | |
| "learning_rate": 0.00010061599466381389, | |
| "loss": 0.0287, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.27880614367112977, | |
| "learning_rate": 0.00010036959829424131, | |
| "loss": 0.0249, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.3435335426279227, | |
| "learning_rate": 0.0001001231996807397, | |
| "loss": 0.0363, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.3166036055942454, | |
| "learning_rate": 9.987680031926032e-05, | |
| "loss": 0.0302, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.2894910254443763, | |
| "learning_rate": 9.96304017057587e-05, | |
| "loss": 0.0285, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.34201749143530447, | |
| "learning_rate": 9.938400533618615e-05, | |
| "loss": 0.0443, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.3740842570087233, | |
| "learning_rate": 9.913761270648015e-05, | |
| "loss": 0.0303, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.3215036425488617, | |
| "learning_rate": 9.889122531255552e-05, | |
| "loss": 0.0383, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.29139970169902235, | |
| "learning_rate": 9.864484465029536e-05, | |
| "loss": 0.0266, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.26997285345644567, | |
| "learning_rate": 9.839847221554175e-05, | |
| "loss": 0.0271, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.37692961834447125, | |
| "learning_rate": 9.815210950408704e-05, | |
| "loss": 0.0409, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 0.34286292595950857, | |
| "learning_rate": 9.790575801166432e-05, | |
| "loss": 0.0373, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 0.3255553751579432, | |
| "learning_rate": 9.765941923393874e-05, | |
| "loss": 0.0272, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.3158220004892916, | |
| "learning_rate": 9.741309466649813e-05, | |
| "loss": 0.0269, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.2736162407266226, | |
| "learning_rate": 9.716678580484411e-05, | |
| "loss": 0.0247, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.337905173595386, | |
| "learning_rate": 9.692049414438299e-05, | |
| "loss": 0.0378, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 0.3258813431481511, | |
| "learning_rate": 9.667422118041651e-05, | |
| "loss": 0.0276, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 0.31627150266086607, | |
| "learning_rate": 9.642796840813308e-05, | |
| "loss": 0.0246, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 0.32285673653685554, | |
| "learning_rate": 9.61817373225983e-05, | |
| "loss": 0.0256, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3021398612416659, | |
| "learning_rate": 9.593552941874635e-05, | |
| "loss": 0.0256, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.2771885920682674, | |
| "learning_rate": 9.568934619137046e-05, | |
| "loss": 0.0236, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.2913089871269692, | |
| "learning_rate": 9.544318913511416e-05, | |
| "loss": 0.0276, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.29943649863489186, | |
| "learning_rate": 9.519705974446207e-05, | |
| "loss": 0.0281, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.28658510456203773, | |
| "learning_rate": 9.495095951373076e-05, | |
| "loss": 0.0222, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.32216521722606845, | |
| "learning_rate": 9.470488993705992e-05, | |
| "loss": 0.0275, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.3534602351211303, | |
| "learning_rate": 9.4458852508403e-05, | |
| "loss": 0.0349, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "eval_blimp_filtered_avg": 0.7164179104477612, | |
| "eval_blimp_filtered_std": 0.005041707326458033, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "eval_blimp_supplement_avg": 0.8297413793103449, | |
| "eval_blimp_supplement_std": 0.016828487437818656, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "eval_vqa_filtered_avg": 0.5, | |
| "eval_vqa_filtered_std": 0.050251890762960605, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "eval_winoground_filtered_avg": 0.66, | |
| "eval_winoground_filtered_std": 0.04760952285695238, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.38483480730152236, | |
| "learning_rate": 9.421284872151836e-05, | |
| "loss": 0.0447, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.2691349907593638, | |
| "learning_rate": 9.396688006996e-05, | |
| "loss": 0.0275, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.30394270832959963, | |
| "learning_rate": 9.372094804706867e-05, | |
| "loss": 0.0245, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.3894950215679374, | |
| "learning_rate": 9.34750541459628e-05, | |
| "loss": 0.0328, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.3066861317584869, | |
| "learning_rate": 9.322919985952926e-05, | |
| "loss": 0.0257, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.2839567453054656, | |
| "learning_rate": 9.298338668041451e-05, | |
| "loss": 0.0301, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 0.3037116435680821, | |
| "learning_rate": 9.27376161010153e-05, | |
| "loss": 0.0221, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 0.3134542478986076, | |
| "learning_rate": 9.249188961346993e-05, | |
| "loss": 0.0338, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.3541455631807929, | |
| "learning_rate": 9.224620870964886e-05, | |
| "loss": 0.0355, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.3115683125480853, | |
| "learning_rate": 9.200057488114585e-05, | |
| "loss": 0.0252, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.27292900060486747, | |
| "learning_rate": 9.175498961926886e-05, | |
| "loss": 0.021, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.28750762243690786, | |
| "learning_rate": 9.150945441503093e-05, | |
| "loss": 0.0262, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.3131812214087709, | |
| "learning_rate": 9.126397075914135e-05, | |
| "loss": 0.032, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.2503186984511136, | |
| "learning_rate": 9.101854014199622e-05, | |
| "loss": 0.0222, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.26585317166475225, | |
| "learning_rate": 9.077316405366981e-05, | |
| "loss": 0.0237, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.30126199934192793, | |
| "learning_rate": 9.052784398390525e-05, | |
| "loss": 0.0294, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.2903130573653526, | |
| "learning_rate": 9.028258142210552e-05, | |
| "loss": 0.0294, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.27174816407170116, | |
| "learning_rate": 9.00373778573246e-05, | |
| "loss": 0.0269, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 0.3629515708975863, | |
| "learning_rate": 8.979223477825814e-05, | |
| "loss": 0.0281, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.3714998200084635, | |
| "learning_rate": 8.954715367323468e-05, | |
| "loss": 0.0331, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.3387130445188704, | |
| "learning_rate": 8.930213603020638e-05, | |
| "loss": 0.0349, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.33356330455180916, | |
| "learning_rate": 8.905718333674013e-05, | |
| "loss": 0.0323, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.3443287065763936, | |
| "learning_rate": 8.881229708000865e-05, | |
| "loss": 0.0314, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.36589381487249345, | |
| "learning_rate": 8.85674787467811e-05, | |
| "loss": 0.0331, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.33686969776224657, | |
| "learning_rate": 8.832272982341439e-05, | |
| "loss": 0.0347, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.2821892144697015, | |
| "learning_rate": 8.80780517958439e-05, | |
| "loss": 0.029, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.3370413152254308, | |
| "learning_rate": 8.783344614957477e-05, | |
| "loss": 0.0309, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 0.27254650378131945, | |
| "learning_rate": 8.758891436967252e-05, | |
| "loss": 0.0206, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 0.344842568070324, | |
| "learning_rate": 8.734445794075428e-05, | |
| "loss": 0.0331, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.3263812096586238, | |
| "learning_rate": 8.710007834697969e-05, | |
| "loss": 0.0323, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.24908057089857247, | |
| "learning_rate": 8.68557770720419e-05, | |
| "loss": 0.024, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.2829456218767608, | |
| "learning_rate": 8.661155559915863e-05, | |
| "loss": 0.0258, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 0.2992177526407328, | |
| "learning_rate": 8.636741541106299e-05, | |
| "loss": 0.0258, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 0.3498647491010859, | |
| "learning_rate": 8.61233579899947e-05, | |
| "loss": 0.0309, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 0.31399097802426235, | |
| "learning_rate": 8.587938481769089e-05, | |
| "loss": 0.0391, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.28430578853728555, | |
| "learning_rate": 8.563549737537719e-05, | |
| "loss": 0.026, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.2523093762154707, | |
| "learning_rate": 8.539169714375885e-05, | |
| "loss": 0.0212, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.29858190065300216, | |
| "learning_rate": 8.514798560301152e-05, | |
| "loss": 0.0339, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.2536866212590354, | |
| "learning_rate": 8.490436423277248e-05, | |
| "loss": 0.0254, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.2924064432366086, | |
| "learning_rate": 8.466083451213144e-05, | |
| "loss": 0.0293, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.2981743418161694, | |
| "learning_rate": 8.441739791962187e-05, | |
| "loss": 0.0258, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.2812452863028169, | |
| "learning_rate": 8.417405593321163e-05, | |
| "loss": 0.0258, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.28479487575175905, | |
| "learning_rate": 8.393081003029431e-05, | |
| "loss": 0.0256, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.34570399590977596, | |
| "learning_rate": 8.368766168768014e-05, | |
| "loss": 0.0325, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.3337375710597649, | |
| "learning_rate": 8.344461238158699e-05, | |
| "loss": 0.0327, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.2986872900474696, | |
| "learning_rate": 8.320166358763159e-05, | |
| "loss": 0.0217, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.28712435596387936, | |
| "learning_rate": 8.295881678082024e-05, | |
| "loss": 0.0256, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.3272426704832391, | |
| "learning_rate": 8.271607343554021e-05, | |
| "loss": 0.0246, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.3092816463373921, | |
| "learning_rate": 8.247343502555053e-05, | |
| "loss": 0.0304, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.3000320928728694, | |
| "learning_rate": 8.223090302397313e-05, | |
| "loss": 0.0246, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.2990018271701567, | |
| "learning_rate": 8.198847890328406e-05, | |
| "loss": 0.0301, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.3674125202368627, | |
| "learning_rate": 8.174616413530418e-05, | |
| "loss": 0.0294, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.32604533269795055, | |
| "learning_rate": 8.150396019119062e-05, | |
| "loss": 0.0345, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 0.283530884706442, | |
| "learning_rate": 8.126186854142752e-05, | |
| "loss": 0.0193, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 0.25164065278234365, | |
| "learning_rate": 8.101989065581743e-05, | |
| "loss": 0.0184, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 0.25607430685124377, | |
| "learning_rate": 8.077802800347205e-05, | |
| "loss": 0.0199, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.2934139007796417, | |
| "learning_rate": 8.053628205280347e-05, | |
| "loss": 0.0234, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.30060563019331127, | |
| "learning_rate": 8.029465427151538e-05, | |
| "loss": 0.0243, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.31584329920715304, | |
| "learning_rate": 8.005314612659393e-05, | |
| "loss": 0.0233, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.3399830490651399, | |
| "learning_rate": 7.9811759084299e-05, | |
| "loss": 0.0303, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.23710075133737013, | |
| "learning_rate": 7.957049461015512e-05, | |
| "loss": 0.0201, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.291414714994268, | |
| "learning_rate": 7.932935416894272e-05, | |
| "loss": 0.0261, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.25516874264704126, | |
| "learning_rate": 7.908833922468927e-05, | |
| "loss": 0.0209, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.3225236014785536, | |
| "learning_rate": 7.884745124066023e-05, | |
| "loss": 0.0293, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 0.33073323071844557, | |
| "learning_rate": 7.860669167935028e-05, | |
| "loss": 0.0312, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 0.2747460833033659, | |
| "learning_rate": 7.836606200247436e-05, | |
| "loss": 0.0249, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.22799207899871587, | |
| "learning_rate": 7.812556367095896e-05, | |
| "loss": 0.0236, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.2957004620104486, | |
| "learning_rate": 7.788519814493304e-05, | |
| "loss": 0.0264, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.28521269666590277, | |
| "learning_rate": 7.764496688371929e-05, | |
| "loss": 0.0279, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 0.24287300349001784, | |
| "learning_rate": 7.740487134582525e-05, | |
| "loss": 0.0182, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 0.3175933387720721, | |
| "learning_rate": 7.716491298893442e-05, | |
| "loss": 0.0339, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.32835585850076665, | |
| "learning_rate": 7.692509326989753e-05, | |
| "loss": 0.0264, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.3273636208785487, | |
| "learning_rate": 7.668541364472346e-05, | |
| "loss": 0.0371, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.32029214184531163, | |
| "learning_rate": 7.644587556857065e-05, | |
| "loss": 0.0266, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.3573218111505275, | |
| "learning_rate": 7.620648049573815e-05, | |
| "loss": 0.0294, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.2639798311874394, | |
| "learning_rate": 7.596722987965669e-05, | |
| "loss": 0.0245, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.3198924686801555, | |
| "learning_rate": 7.572812517288018e-05, | |
| "loss": 0.0222, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.3153205405405442, | |
| "learning_rate": 7.548916782707642e-05, | |
| "loss": 0.0322, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.26231028932152634, | |
| "learning_rate": 7.525035929301877e-05, | |
| "loss": 0.0302, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.2688781021142035, | |
| "learning_rate": 7.50117010205769e-05, | |
| "loss": 0.0251, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.26830771551596, | |
| "learning_rate": 7.477319445870845e-05, | |
| "loss": 0.0226, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.30176145649756575, | |
| "learning_rate": 7.453484105544976e-05, | |
| "loss": 0.0331, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.3115716482147434, | |
| "learning_rate": 7.429664225790743e-05, | |
| "loss": 0.0233, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.31100418685578723, | |
| "learning_rate": 7.405859951224933e-05, | |
| "loss": 0.0284, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.2588637124537901, | |
| "learning_rate": 7.382071426369597e-05, | |
| "loss": 0.0233, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 0.31963780634052935, | |
| "learning_rate": 7.358298795651165e-05, | |
| "loss": 0.0201, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 0.3357482742804347, | |
| "learning_rate": 7.33454220339956e-05, | |
| "loss": 0.0234, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.31241871424393247, | |
| "learning_rate": 7.310801793847344e-05, | |
| "loss": 0.0352, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.34133470506967756, | |
| "learning_rate": 7.287077711128823e-05, | |
| "loss": 0.0237, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.14462364414170686, | |
| "learning_rate": 7.263370099279172e-05, | |
| "loss": 0.0101, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 0.139685520904427, | |
| "learning_rate": 7.239679102233582e-05, | |
| "loss": 0.008, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 0.15075598570023135, | |
| "learning_rate": 7.21600486382636e-05, | |
| "loss": 0.009, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.09798572261973783, | |
| "learning_rate": 7.192347527790073e-05, | |
| "loss": 0.0047, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.17274723978571413, | |
| "learning_rate": 7.168707237754658e-05, | |
| "loss": 0.0101, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.14385963722120923, | |
| "learning_rate": 7.14508413724658e-05, | |
| "loss": 0.0087, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.1417300685940669, | |
| "learning_rate": 7.121478369687926e-05, | |
| "loss": 0.005, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.13023296035033566, | |
| "learning_rate": 7.097890078395553e-05, | |
| "loss": 0.0062, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.15199661343358178, | |
| "learning_rate": 7.074319406580224e-05, | |
| "loss": 0.0077, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.26597073732306437, | |
| "learning_rate": 7.050766497345714e-05, | |
| "loss": 0.0203, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.16538027691964055, | |
| "learning_rate": 7.027231493687974e-05, | |
| "loss": 0.0111, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "eval_blimp_filtered_avg": 0.7185074626865672, | |
| "eval_blimp_filtered_std": 0.005025927855759427, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "eval_blimp_supplement_avg": 0.8297413793103449, | |
| "eval_blimp_supplement_std": 0.016768829882349248, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "eval_vqa_filtered_avg": 0.49, | |
| "eval_vqa_filtered_std": 0.05024183937956912, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "eval_winoground_filtered_avg": 0.67, | |
| "eval_winoground_filtered_std": 0.04725815626252606, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.21979041950155864, | |
| "learning_rate": 7.003714538494233e-05, | |
| "loss": 0.0087, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.18913233160499004, | |
| "learning_rate": 6.980215774542147e-05, | |
| "loss": 0.0096, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.1282988164073535, | |
| "learning_rate": 6.95673534449893e-05, | |
| "loss": 0.0068, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 0.21752972603923165, | |
| "learning_rate": 6.933273390920478e-05, | |
| "loss": 0.0104, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 0.1967481497472081, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.01, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 0.18179145828619866, | |
| "learning_rate": 6.886405482819756e-05, | |
| "loss": 0.0093, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.17276847874184295, | |
| "learning_rate": 6.862999812844953e-05, | |
| "loss": 0.0092, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.1342024775407428, | |
| "learning_rate": 6.839613188428126e-05, | |
| "loss": 0.0065, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.14878102990268274, | |
| "learning_rate": 6.81624575155566e-05, | |
| "loss": 0.0063, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.16403017563876973, | |
| "learning_rate": 6.792897644097451e-05, | |
| "loss": 0.0087, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.19801520787974072, | |
| "learning_rate": 6.769569007806027e-05, | |
| "loss": 0.0093, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.20572707361042517, | |
| "learning_rate": 6.746259984315717e-05, | |
| "loss": 0.011, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.13697981744457283, | |
| "learning_rate": 6.722970715141763e-05, | |
| "loss": 0.0046, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.19205653189851232, | |
| "learning_rate": 6.699701341679488e-05, | |
| "loss": 0.0092, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.1918525797269716, | |
| "learning_rate": 6.676452005203406e-05, | |
| "loss": 0.0098, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.2248320215810751, | |
| "learning_rate": 6.653222846866389e-05, | |
| "loss": 0.0109, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.15114455523839437, | |
| "learning_rate": 6.630014007698807e-05, | |
| "loss": 0.0067, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.13233475079909107, | |
| "learning_rate": 6.606825628607654e-05, | |
| "loss": 0.0071, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.13146130811845824, | |
| "learning_rate": 6.583657850375723e-05, | |
| "loss": 0.0059, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.1857326870966165, | |
| "learning_rate": 6.560510813660719e-05, | |
| "loss": 0.0036, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.2298298643113706, | |
| "learning_rate": 6.537384658994428e-05, | |
| "loss": 0.0086, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.14279703425417056, | |
| "learning_rate": 6.51427952678185e-05, | |
| "loss": 0.0085, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.10832021907148229, | |
| "learning_rate": 6.491195557300353e-05, | |
| "loss": 0.0058, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 0.11823319091468878, | |
| "learning_rate": 6.468132890698829e-05, | |
| "loss": 0.0062, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.18897801838669076, | |
| "learning_rate": 6.44509166699682e-05, | |
| "loss": 0.0092, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.25020039121328774, | |
| "learning_rate": 6.422072026083697e-05, | |
| "loss": 0.0137, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.20775807254077025, | |
| "learning_rate": 6.399074107717782e-05, | |
| "loss": 0.0093, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.20035420751641417, | |
| "learning_rate": 6.376098051525529e-05, | |
| "loss": 0.008, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.1813508652414028, | |
| "learning_rate": 6.35314399700065e-05, | |
| "loss": 0.0046, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.17007027062456834, | |
| "learning_rate": 6.33021208350328e-05, | |
| "loss": 0.0073, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.19015476370218232, | |
| "learning_rate": 6.307302450259136e-05, | |
| "loss": 0.0106, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.16589400476299693, | |
| "learning_rate": 6.284415236358653e-05, | |
| "loss": 0.0055, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.16209318707015577, | |
| "learning_rate": 6.261550580756175e-05, | |
| "loss": 0.005, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.18839393523384632, | |
| "learning_rate": 6.238708622269065e-05, | |
| "loss": 0.0089, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.16926357657366423, | |
| "learning_rate": 6.215889499576898e-05, | |
| "loss": 0.0071, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.1427288948192287, | |
| "learning_rate": 6.193093351220605e-05, | |
| "loss": 0.0071, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.0823573599543454, | |
| "learning_rate": 6.170320315601628e-05, | |
| "loss": 0.0028, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 0.17793420603102786, | |
| "learning_rate": 6.147570530981099e-05, | |
| "loss": 0.0073, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 0.2613656409512038, | |
| "learning_rate": 6.12484413547897e-05, | |
| "loss": 0.0099, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "grad_norm": 0.10748497227166323, | |
| "learning_rate": 6.102141267073207e-05, | |
| "loss": 0.0056, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3115342005600293, | |
| "learning_rate": 6.0794620635989244e-05, | |
| "loss": 0.0202, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.17604863819079872, | |
| "learning_rate": 6.056806662747572e-05, | |
| "loss": 0.0058, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 0.09017231755285983, | |
| "learning_rate": 6.034175202066077e-05, | |
| "loss": 0.0031, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 0.15807640643888599, | |
| "learning_rate": 6.011567818956021e-05, | |
| "loss": 0.0083, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "grad_norm": 0.1322797473736452, | |
| "learning_rate": 5.988984650672813e-05, | |
| "loss": 0.0054, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.1646419558157819, | |
| "learning_rate": 5.96642583432484e-05, | |
| "loss": 0.0088, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.15547257435043996, | |
| "learning_rate": 5.943891506872645e-05, | |
| "loss": 0.0074, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.19268286944193538, | |
| "learning_rate": 5.921381805128088e-05, | |
| "loss": 0.0125, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.16001416685280997, | |
| "learning_rate": 5.898896865753522e-05, | |
| "loss": 0.0102, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.14560946392731133, | |
| "learning_rate": 5.876436825260967e-05, | |
| "loss": 0.0058, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 0.10597553502655095, | |
| "learning_rate": 5.854001820011265e-05, | |
| "loss": 0.0042, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 0.1477597061504793, | |
| "learning_rate": 5.831591986213274e-05, | |
| "loss": 0.0115, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 0.11589346222710697, | |
| "learning_rate": 5.809207459923016e-05, | |
| "loss": 0.0052, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.11233318637657551, | |
| "learning_rate": 5.786848377042875e-05, | |
| "loss": 0.0047, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.14038827344355653, | |
| "learning_rate": 5.764514873320761e-05, | |
| "loss": 0.0056, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.13670003396305133, | |
| "learning_rate": 5.7422070843492734e-05, | |
| "loss": 0.007, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 0.10138719221994133, | |
| "learning_rate": 5.719925145564913e-05, | |
| "loss": 0.005, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 0.10712806624372671, | |
| "learning_rate": 5.697669192247215e-05, | |
| "loss": 0.0032, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 0.2867594004405392, | |
| "learning_rate": 5.675439359517962e-05, | |
| "loss": 0.007, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 0.15537957888603157, | |
| "learning_rate": 5.6532357823403517e-05, | |
| "loss": 0.0059, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 0.13908151142988748, | |
| "learning_rate": 5.63105859551817e-05, | |
| "loss": 0.0055, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 0.1670760871245682, | |
| "learning_rate": 5.608907933694994e-05, | |
| "loss": 0.007, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 3.28, | |
| "grad_norm": 0.15128774900728617, | |
| "learning_rate": 5.586783931353338e-05, | |
| "loss": 0.0071, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.118351563993401, | |
| "learning_rate": 5.56468672281388e-05, | |
| "loss": 0.0052, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.26632088069900783, | |
| "learning_rate": 5.542616442234618e-05, | |
| "loss": 0.0096, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.15185019444874828, | |
| "learning_rate": 5.5205732236100635e-05, | |
| "loss": 0.0065, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.1448611480489451, | |
| "learning_rate": 5.498557200770429e-05, | |
| "loss": 0.0065, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.16431150027248104, | |
| "learning_rate": 5.476568507380815e-05, | |
| "loss": 0.0066, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.11908895770673851, | |
| "learning_rate": 5.454607276940389e-05, | |
| "loss": 0.0032, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.1443718220854692, | |
| "learning_rate": 5.4326736427815946e-05, | |
| "loss": 0.0053, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.14871442319667327, | |
| "learning_rate": 5.410767738069328e-05, | |
| "loss": 0.0059, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 0.18082767972242653, | |
| "learning_rate": 5.388889695800129e-05, | |
| "loss": 0.0068, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 0.1648249037129662, | |
| "learning_rate": 5.3670396488013854e-05, | |
| "loss": 0.0067, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 0.12517423729992644, | |
| "learning_rate": 5.345217729730501e-05, | |
| "loss": 0.0048, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.1464921521793365, | |
| "learning_rate": 5.3234240710741337e-05, | |
| "loss": 0.0068, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.17291303731043475, | |
| "learning_rate": 5.301658805147338e-05, | |
| "loss": 0.0096, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.1280298983910598, | |
| "learning_rate": 5.279922064092808e-05, | |
| "loss": 0.0034, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 0.12917298155423965, | |
| "learning_rate": 5.25821397988005e-05, | |
| "loss": 0.0043, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 0.16991548955604624, | |
| "learning_rate": 5.236534684304575e-05, | |
| "loss": 0.0036, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.1341181151945374, | |
| "learning_rate": 5.214884308987136e-05, | |
| "loss": 0.0046, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.07618404508173873, | |
| "learning_rate": 5.193262985372879e-05, | |
| "loss": 0.0027, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.08880977813478996, | |
| "learning_rate": 5.171670844730581e-05, | |
| "loss": 0.0036, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.13216208460528805, | |
| "learning_rate": 5.150108018151845e-05, | |
| "loss": 0.0042, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.11790812970039133, | |
| "learning_rate": 5.128574636550283e-05, | |
| "loss": 0.0035, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 0.11136507705857467, | |
| "learning_rate": 5.107070830660765e-05, | |
| "loss": 0.0032, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 0.1625804231101919, | |
| "learning_rate": 5.0855967310385776e-05, | |
| "loss": 0.0035, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 0.16854543758219576, | |
| "learning_rate": 5.064152468058661e-05, | |
| "loss": 0.013, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 0.1904583816627798, | |
| "learning_rate": 5.0427381719148115e-05, | |
| "loss": 0.0058, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 0.1407250892066329, | |
| "learning_rate": 5.021353972618877e-05, | |
| "loss": 0.0048, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 0.09332762885645429, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.0033, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.1282598573789377, | |
| "learning_rate": 4.978676383703792e-05, | |
| "loss": 0.0051, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.12527309494209085, | |
| "learning_rate": 4.957383253191567e-05, | |
| "loss": 0.0039, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.1474717370503847, | |
| "learning_rate": 4.9361207377395526e-05, | |
| "loss": 0.0056, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.17216191154710148, | |
| "learning_rate": 4.914888966438107e-05, | |
| "loss": 0.0061, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.11969436074946377, | |
| "learning_rate": 4.893688068190932e-05, | |
| "loss": 0.0045, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 0.16780413116289464, | |
| "learning_rate": 4.872518171714285e-05, | |
| "loss": 0.0067, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 0.09403083909006464, | |
| "learning_rate": 4.8513794055362094e-05, | |
| "loss": 0.0033, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 0.09334786241295161, | |
| "learning_rate": 4.8302718979957465e-05, | |
| "loss": 0.0025, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 0.12070714502867205, | |
| "learning_rate": 4.809195777242157e-05, | |
| "loss": 0.0044, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 0.0724744178923367, | |
| "learning_rate": 4.7881511712341484e-05, | |
| "loss": 0.0029, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "eval_blimp_filtered_avg": 0.7164179104477612, | |
| "eval_blimp_filtered_std": 0.0050468818783488715, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "eval_blimp_supplement_avg": 0.8318965517241379, | |
| "eval_blimp_supplement_std": 0.016663496994065188, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "eval_vqa_filtered_avg": 0.48, | |
| "eval_vqa_filtered_std": 0.05021167315686779, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "eval_winoground_filtered_avg": 0.68, | |
| "eval_winoground_filtered_std": 0.046882617226215034, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.12264611037048552, | |
| "learning_rate": 4.7671382077390923e-05, | |
| "loss": 0.0049, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.1447159721309767, | |
| "learning_rate": 4.746157014332242e-05, | |
| "loss": 0.0044, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 0.16375949938502024, | |
| "learning_rate": 4.7252077183959766e-05, | |
| "loss": 0.0065, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.10516324597971437, | |
| "learning_rate": 4.704290447119013e-05, | |
| "loss": 0.0043, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.08499567671503878, | |
| "learning_rate": 4.683405327495638e-05, | |
| "loss": 0.0033, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.24460270399320144, | |
| "learning_rate": 4.6625524863249435e-05, | |
| "loss": 0.0073, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.12925387617205644, | |
| "learning_rate": 4.6417320502100316e-05, | |
| "loss": 0.0044, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.09078563973748639, | |
| "learning_rate": 4.6209441455572934e-05, | |
| "loss": 0.0024, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 0.15653536427369144, | |
| "learning_rate": 4.600188898575585e-05, | |
| "loss": 0.0069, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 0.2617481572374967, | |
| "learning_rate": 4.5794664352755055e-05, | |
| "loss": 0.0078, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 0.16686255253211194, | |
| "learning_rate": 4.558776881468616e-05, | |
| "loss": 0.0063, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 0.14235218284923637, | |
| "learning_rate": 4.538120362766659e-05, | |
| "loss": 0.0043, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 0.2061352593588118, | |
| "learning_rate": 4.5174970045808373e-05, | |
| "loss": 0.0083, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 0.09141171412747422, | |
| "learning_rate": 4.496906932121006e-05, | |
| "loss": 0.0038, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 0.14581164354276802, | |
| "learning_rate": 4.476350270394942e-05, | |
| "loss": 0.0054, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 0.1909792639108279, | |
| "learning_rate": 4.4558271442075817e-05, | |
| "loss": 0.0058, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.16782053974713387, | |
| "learning_rate": 4.435337678160244e-05, | |
| "loss": 0.0053, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.1679302657777338, | |
| "learning_rate": 4.414881996649909e-05, | |
| "loss": 0.006, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.11555008314937329, | |
| "learning_rate": 4.394460223868422e-05, | |
| "loss": 0.0038, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.13987569584452733, | |
| "learning_rate": 4.374072483801769e-05, | |
| "loss": 0.0057, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.15472739669003643, | |
| "learning_rate": 4.353718900229315e-05, | |
| "loss": 0.0047, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.11860213300146173, | |
| "learning_rate": 4.333399596723054e-05, | |
| "loss": 0.0041, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.12771536456303176, | |
| "learning_rate": 4.313114696646844e-05, | |
| "loss": 0.0051, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.1786915439582824, | |
| "learning_rate": 4.2928643231556844e-05, | |
| "loss": 0.0059, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.24120215565132727, | |
| "learning_rate": 4.272648599194948e-05, | |
| "loss": 0.0046, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.14109245430082046, | |
| "learning_rate": 4.2524676474996436e-05, | |
| "loss": 0.0045, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.08773485394717533, | |
| "learning_rate": 4.232321590593672e-05, | |
| "loss": 0.0026, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 0.11777262879064121, | |
| "learning_rate": 4.212210550789066e-05, | |
| "loss": 0.0039, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 0.15303150859624717, | |
| "learning_rate": 4.192134650185271e-05, | |
| "loss": 0.0049, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 0.17847021642871766, | |
| "learning_rate": 4.172094010668391e-05, | |
| "loss": 0.0122, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 0.12872303754581352, | |
| "learning_rate": 4.1520887539104516e-05, | |
| "loss": 0.0034, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 0.09422413881251451, | |
| "learning_rate": 4.132119001368658e-05, | |
| "loss": 0.0032, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.11765015102315801, | |
| "learning_rate": 4.112184874284655e-05, | |
| "loss": 0.004, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.1816646133957582, | |
| "learning_rate": 4.092286493683812e-05, | |
| "loss": 0.0054, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.08821639897774167, | |
| "learning_rate": 4.072423980374452e-05, | |
| "loss": 0.0027, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.109204348627212, | |
| "learning_rate": 4.052597454947151e-05, | |
| "loss": 0.0037, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.07762991091541192, | |
| "learning_rate": 4.0328070377739936e-05, | |
| "loss": 0.0048, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 0.12303959337300706, | |
| "learning_rate": 4.0130528490078255e-05, | |
| "loss": 0.0054, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 0.14938977746099588, | |
| "learning_rate": 3.993335008581569e-05, | |
| "loss": 0.0054, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 0.10793183299648912, | |
| "learning_rate": 3.973653636207437e-05, | |
| "loss": 0.0051, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 0.07679917764318481, | |
| "learning_rate": 3.954008851376252e-05, | |
| "loss": 0.0021, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 0.20216658828207906, | |
| "learning_rate": 3.934400773356702e-05, | |
| "loss": 0.0082, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 0.17081936190481015, | |
| "learning_rate": 3.914829521194606e-05, | |
| "loss": 0.0047, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 0.1681844318722247, | |
| "learning_rate": 3.895295213712227e-05, | |
| "loss": 0.0053, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 0.11660462512191513, | |
| "learning_rate": 3.875797969507502e-05, | |
| "loss": 0.0068, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.10899837547854943, | |
| "learning_rate": 3.8563379069533626e-05, | |
| "loss": 0.004, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.11737468807394794, | |
| "learning_rate": 3.836915144196995e-05, | |
| "loss": 0.0043, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.17750707113239161, | |
| "learning_rate": 3.8175297991591316e-05, | |
| "loss": 0.0059, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 0.08434750541376605, | |
| "learning_rate": 3.7981819895333336e-05, | |
| "loss": 0.0028, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 0.08977864719318272, | |
| "learning_rate": 3.778871832785262e-05, | |
| "loss": 0.0025, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.26117422627907305, | |
| "learning_rate": 3.759599446151994e-05, | |
| "loss": 0.0058, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.1423125205304331, | |
| "learning_rate": 3.740364946641284e-05, | |
| "loss": 0.005, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.048844504093019145, | |
| "learning_rate": 3.721168451030868e-05, | |
| "loss": 0.0013, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 0.12750445172306626, | |
| "learning_rate": 3.702010075867748e-05, | |
| "loss": 0.0039, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 0.1363488868484316, | |
| "learning_rate": 3.682889937467493e-05, | |
| "loss": 0.0054, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 0.10861963297037971, | |
| "learning_rate": 3.6638081519135115e-05, | |
| "loss": 0.0025, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.07423146974751119, | |
| "learning_rate": 3.6447648350563767e-05, | |
| "loss": 0.002, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.1473106789075466, | |
| "learning_rate": 3.6257601025131026e-05, | |
| "loss": 0.0048, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.20756086874083257, | |
| "learning_rate": 3.6067940696664484e-05, | |
| "loss": 0.0044, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.12432385236166127, | |
| "learning_rate": 3.587866851664219e-05, | |
| "loss": 0.0042, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.1555267425126639, | |
| "learning_rate": 3.568978563418551e-05, | |
| "loss": 0.004, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 0.10864709636610591, | |
| "learning_rate": 3.5501293196052544e-05, | |
| "loss": 0.003, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 0.10844528424446415, | |
| "learning_rate": 3.531319234663063e-05, | |
| "loss": 0.003, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.18756992146351528, | |
| "learning_rate": 3.512548422792983e-05, | |
| "loss": 0.0039, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.16307530028534215, | |
| "learning_rate": 3.493816997957582e-05, | |
| "loss": 0.0036, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.21835831614587714, | |
| "learning_rate": 3.4751250738802835e-05, | |
| "loss": 0.0057, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.12509619886030035, | |
| "learning_rate": 3.456472764044718e-05, | |
| "loss": 0.0037, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.1878620997227208, | |
| "learning_rate": 3.4378601816939824e-05, | |
| "loss": 0.0044, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.20067996408678637, | |
| "learning_rate": 3.4192874398299915e-05, | |
| "loss": 0.0074, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 0.09941229298151419, | |
| "learning_rate": 3.400754651212776e-05, | |
| "loss": 0.0038, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 0.13212089570245356, | |
| "learning_rate": 3.382261928359791e-05, | |
| "loss": 0.0047, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.08933394980267556, | |
| "learning_rate": 3.36380938354526e-05, | |
| "loss": 0.0039, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.16729919593124423, | |
| "learning_rate": 3.3453971287994545e-05, | |
| "loss": 0.0079, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.15697707244755352, | |
| "learning_rate": 3.3270252759080476e-05, | |
| "loss": 0.0052, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.09194443755715154, | |
| "learning_rate": 3.308693936411421e-05, | |
| "loss": 0.0024, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.04543214336381121, | |
| "learning_rate": 3.290403221603976e-05, | |
| "loss": 0.0014, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.13598936920247842, | |
| "learning_rate": 3.2721532425334934e-05, | |
| "loss": 0.003, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 0.12519394602390674, | |
| "learning_rate": 3.253944110000415e-05, | |
| "loss": 0.0044, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 0.10159326068686843, | |
| "learning_rate": 3.235775934557204e-05, | |
| "loss": 0.0037, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 0.09206667100999626, | |
| "learning_rate": 3.2176488265076596e-05, | |
| "loss": 0.0023, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 0.13180534984976824, | |
| "learning_rate": 3.199562895906252e-05, | |
| "loss": 0.0059, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 0.1164365775293987, | |
| "learning_rate": 3.1815182525574495e-05, | |
| "loss": 0.0039, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 0.09037223834103687, | |
| "learning_rate": 3.163515006015052e-05, | |
| "loss": 0.0031, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 0.1243136017465312, | |
| "learning_rate": 3.1455532655815346e-05, | |
| "loss": 0.0023, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.06736081989652704, | |
| "learning_rate": 3.1276331403073735e-05, | |
| "loss": 0.0018, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.17307715618033684, | |
| "learning_rate": 3.10975473899039e-05, | |
| "loss": 0.005, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.06048816505854376, | |
| "learning_rate": 3.09191817017509e-05, | |
| "loss": 0.0021, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.12504484352987397, | |
| "learning_rate": 3.074123542152001e-05, | |
| "loss": 0.0032, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.06833118623733686, | |
| "learning_rate": 3.056370962957014e-05, | |
| "loss": 0.0021, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.1108735830960286, | |
| "learning_rate": 3.0386605403707346e-05, | |
| "loss": 0.0034, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 0.09939702792131645, | |
| "learning_rate": 3.020992381917823e-05, | |
| "loss": 0.0029, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 0.0991719206432376, | |
| "learning_rate": 3.0033665948663448e-05, | |
| "loss": 0.0033, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.12662565643384302, | |
| "learning_rate": 2.9857832862271183e-05, | |
| "loss": 0.004, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.20781158845550377, | |
| "learning_rate": 2.968242562753051e-05, | |
| "loss": 0.0058, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.14694222050004627, | |
| "learning_rate": 2.9507445309385294e-05, | |
| "loss": 0.0115, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 0.10808293020495671, | |
| "learning_rate": 2.9332892970187255e-05, | |
| "loss": 0.0043, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 0.12048492965543693, | |
| "learning_rate": 2.915876966968978e-05, | |
| "loss": 0.0029, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 0.09225586697033628, | |
| "learning_rate": 2.8985076465041582e-05, | |
| "loss": 0.0023, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.16415643694268406, | |
| "learning_rate": 2.8811814410779957e-05, | |
| "loss": 0.0055, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.11278583943799678, | |
| "learning_rate": 2.8638984558824777e-05, | |
| "loss": 0.0039, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "eval_blimp_filtered_avg": 0.716865671641791, | |
| "eval_blimp_filtered_std": 0.00502778551816782, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "eval_blimp_supplement_avg": 0.8362068965517241, | |
| "eval_blimp_supplement_std": 0.016430816592740906, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "eval_vqa_filtered_avg": 0.48, | |
| "eval_vqa_filtered_std": 0.05021167315686779, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "eval_winoground_filtered_avg": 0.7, | |
| "eval_winoground_filtered_std": 0.046056618647183814, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.19790160401886048, | |
| "learning_rate": 2.8466587958471713e-05, | |
| "loss": 0.0079, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.15241865463221207, | |
| "learning_rate": 2.8294625656386153e-05, | |
| "loss": 0.0033, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 0.15486582212720096, | |
| "learning_rate": 2.812309869659675e-05, | |
| "loss": 0.0049, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.1403453911129104, | |
| "learning_rate": 2.7952008120489005e-05, | |
| "loss": 0.0035, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.2066076788884496, | |
| "learning_rate": 2.7781354966799078e-05, | |
| "loss": 0.0069, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 0.14517188108322096, | |
| "learning_rate": 2.7611140271607417e-05, | |
| "loss": 0.0038, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 0.19311241295438267, | |
| "learning_rate": 2.744136506833247e-05, | |
| "loss": 0.0046, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 0.15261908108180047, | |
| "learning_rate": 2.7272030387724423e-05, | |
| "loss": 0.0051, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.148487491757181, | |
| "learning_rate": 2.7103137257858868e-05, | |
| "loss": 0.0047, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2545086755837555, | |
| "learning_rate": 2.6934686704130696e-05, | |
| "loss": 0.0078, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2155734204394312, | |
| "learning_rate": 2.6766679749247793e-05, | |
| "loss": 0.0053, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.11633424479852102, | |
| "learning_rate": 2.6599117413224817e-05, | |
| "loss": 0.0026, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.16185230460669028, | |
| "learning_rate": 2.6432000713377027e-05, | |
| "loss": 0.006, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 0.06116976817258535, | |
| "learning_rate": 2.6265330664314157e-05, | |
| "loss": 0.002, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 0.1467173492208502, | |
| "learning_rate": 2.6099108277934103e-05, | |
| "loss": 0.0045, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 0.06626573294126052, | |
| "learning_rate": 2.5933334563416976e-05, | |
| "loss": 0.002, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.12063079499322236, | |
| "learning_rate": 2.5768010527218845e-05, | |
| "loss": 0.0027, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.10686596344627741, | |
| "learning_rate": 2.5603137173065674e-05, | |
| "loss": 0.0043, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.11272599358320497, | |
| "learning_rate": 2.543871550194723e-05, | |
| "loss": 0.0027, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 0.1718157414634651, | |
| "learning_rate": 2.527474651211089e-05, | |
| "loss": 0.0041, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 0.11357917775880114, | |
| "learning_rate": 2.5111231199055896e-05, | |
| "loss": 0.0038, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 0.07702764510610083, | |
| "learning_rate": 2.494817055552686e-05, | |
| "loss": 0.0025, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 0.1682497107538642, | |
| "learning_rate": 2.4785565571508118e-05, | |
| "loss": 0.007, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 0.06959209717037654, | |
| "learning_rate": 2.462341723421758e-05, | |
| "loss": 0.0021, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.1216172792972272, | |
| "learning_rate": 2.4461726528100615e-05, | |
| "loss": 0.0034, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.1513509657366934, | |
| "learning_rate": 2.4300494434824373e-05, | |
| "loss": 0.004, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.0865648646398974, | |
| "learning_rate": 2.4139721933271465e-05, | |
| "loss": 0.0024, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 0.08556831061548267, | |
| "learning_rate": 2.3979409999534298e-05, | |
| "loss": 0.0023, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 0.1641906844837052, | |
| "learning_rate": 2.381955960690906e-05, | |
| "loss": 0.0075, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.11390717886890653, | |
| "learning_rate": 2.36601717258897e-05, | |
| "loss": 0.003, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.11731050753260445, | |
| "learning_rate": 2.35012473241623e-05, | |
| "loss": 0.0028, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 0.16970688657951408, | |
| "learning_rate": 2.3342787366598872e-05, | |
| "loss": 0.0043, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 0.11089863057017167, | |
| "learning_rate": 2.3184792815251766e-05, | |
| "loss": 0.0032, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 0.046462119059264825, | |
| "learning_rate": 2.302726462934769e-05, | |
| "loss": 0.0018, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.11684974119182631, | |
| "learning_rate": 2.2870203765281926e-05, | |
| "loss": 0.0027, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.16444221243072213, | |
| "learning_rate": 2.2713611176612582e-05, | |
| "loss": 0.0046, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.1818375194889315, | |
| "learning_rate": 2.25574878140546e-05, | |
| "loss": 0.0051, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.11786328596160862, | |
| "learning_rate": 2.240183462547427e-05, | |
| "loss": 0.0024, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.08316727437427408, | |
| "learning_rate": 2.224665255588325e-05, | |
| "loss": 0.0031, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.07270479819118501, | |
| "learning_rate": 2.2091942547432955e-05, | |
| "loss": 0.0027, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.05459676934354277, | |
| "learning_rate": 2.193770553940876e-05, | |
| "loss": 0.0016, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.11956646260392087, | |
| "learning_rate": 2.1783942468224382e-05, | |
| "loss": 0.0018, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 0.08747129822087717, | |
| "learning_rate": 2.163065426741603e-05, | |
| "loss": 0.0025, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 0.1777877170154106, | |
| "learning_rate": 2.147784186763696e-05, | |
| "loss": 0.0043, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 0.08218483908984636, | |
| "learning_rate": 2.132550619665168e-05, | |
| "loss": 0.0018, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.13582850245610714, | |
| "learning_rate": 2.117364817933033e-05, | |
| "loss": 0.0047, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.3137890453832542, | |
| "learning_rate": 2.1022268737643138e-05, | |
| "loss": 0.0065, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.14756824814987923, | |
| "learning_rate": 2.08713687906547e-05, | |
| "loss": 0.0046, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 0.12543766199519707, | |
| "learning_rate": 2.0720949254518517e-05, | |
| "loss": 0.0033, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 0.1272943584540691, | |
| "learning_rate": 2.05710110424714e-05, | |
| "loss": 0.0046, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.18454830615921256, | |
| "learning_rate": 2.0421555064827878e-05, | |
| "loss": 0.0072, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.1537809820642007, | |
| "learning_rate": 2.0272582228974792e-05, | |
| "loss": 0.0043, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.05312631536933881, | |
| "learning_rate": 2.012409343936551e-05, | |
| "loss": 0.0012, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 0.04447185122963781, | |
| "learning_rate": 1.9976089597514903e-05, | |
| "loss": 0.0012, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "grad_norm": 0.026337967004294745, | |
| "learning_rate": 1.982857160199334e-05, | |
| "loss": 0.0007, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.026301866799899334, | |
| "learning_rate": 1.9681540348421623e-05, | |
| "loss": 0.001, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.05436305337809422, | |
| "learning_rate": 1.9534996729465426e-05, | |
| "loss": 0.0014, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.05631442649400675, | |
| "learning_rate": 1.938894163482974e-05, | |
| "loss": 0.0015, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.10283079607735844, | |
| "learning_rate": 1.9243375951253796e-05, | |
| "loss": 0.0023, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.050174636822106565, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.0011, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.018160762498011917, | |
| "learning_rate": 1.895371634937525e-05, | |
| "loss": 0.0008, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 0.053625104843317006, | |
| "learning_rate": 1.880962418967279e-05, | |
| "loss": 0.0017, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 0.0367115996959884, | |
| "learning_rate": 1.8666024958219408e-05, | |
| "loss": 0.0013, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 0.01705265689595668, | |
| "learning_rate": 1.852291952684414e-05, | |
| "loss": 0.0006, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 0.029645947416295275, | |
| "learning_rate": 1.8380308764377842e-05, | |
| "loss": 0.001, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 0.02616977130242847, | |
| "learning_rate": 1.8238193536648195e-05, | |
| "loss": 0.001, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.029725522847751006, | |
| "learning_rate": 1.8096574706474333e-05, | |
| "loss": 0.0011, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.046999003971840925, | |
| "learning_rate": 1.795545313366166e-05, | |
| "loss": 0.0016, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.0639334214169769, | |
| "learning_rate": 1.7814829674996592e-05, | |
| "loss": 0.002, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 0.0367804832140112, | |
| "learning_rate": 1.767470518424129e-05, | |
| "loss": 0.0011, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "grad_norm": 0.029348342947469082, | |
| "learning_rate": 1.7535080512128632e-05, | |
| "loss": 0.0008, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.0466731334600783, | |
| "learning_rate": 1.7395956506356937e-05, | |
| "loss": 0.0015, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.027508272565636798, | |
| "learning_rate": 1.7257334011584847e-05, | |
| "loss": 0.0009, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.035628751171613196, | |
| "learning_rate": 1.7119213869426197e-05, | |
| "loss": 0.001, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 0.06971830663806321, | |
| "learning_rate": 1.6981596918444953e-05, | |
| "loss": 0.0012, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 0.021694267386634027, | |
| "learning_rate": 1.684448399414994e-05, | |
| "loss": 0.0009, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.02007528840430175, | |
| "learning_rate": 1.6707875928990058e-05, | |
| "loss": 0.0008, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.04559222251809475, | |
| "learning_rate": 1.6571773552349e-05, | |
| "loss": 0.0011, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.03837022867264384, | |
| "learning_rate": 1.6436177690540243e-05, | |
| "loss": 0.0013, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 0.0596087088326944, | |
| "learning_rate": 1.630108916680223e-05, | |
| "loss": 0.0017, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 0.017751318308507253, | |
| "learning_rate": 1.6166508801293013e-05, | |
| "loss": 0.0006, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 0.026619692717408015, | |
| "learning_rate": 1.603243741108571e-05, | |
| "loss": 0.0013, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 0.020963517608948525, | |
| "learning_rate": 1.5898875810163137e-05, | |
| "loss": 0.0007, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "grad_norm": 0.027136972648439773, | |
| "learning_rate": 1.5765824809413056e-05, | |
| "loss": 0.0009, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.05016619274243327, | |
| "learning_rate": 1.5633285216623385e-05, | |
| "loss": 0.0013, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.029039341510529837, | |
| "learning_rate": 1.5501257836476978e-05, | |
| "loss": 0.0009, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.07303975658197855, | |
| "learning_rate": 1.5369743470547027e-05, | |
| "loss": 0.0019, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 0.03350400586584435, | |
| "learning_rate": 1.5238742917292015e-05, | |
| "loss": 0.0009, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 0.08029935115169112, | |
| "learning_rate": 1.5108256972050972e-05, | |
| "loss": 0.0019, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 0.04476861542864655, | |
| "learning_rate": 1.4978286427038601e-05, | |
| "loss": 0.0014, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 0.0692481373749192, | |
| "learning_rate": 1.4848832071340423e-05, | |
| "loss": 0.0014, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 0.0459173556561311, | |
| "learning_rate": 1.4719894690908098e-05, | |
| "loss": 0.0013, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.05681694784227542, | |
| "learning_rate": 1.4591475068554572e-05, | |
| "loss": 0.0017, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.03490009549853983, | |
| "learning_rate": 1.4463573983949341e-05, | |
| "loss": 0.001, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.08256537486908391, | |
| "learning_rate": 1.4336192213613742e-05, | |
| "loss": 0.0029, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.03304929283596535, | |
| "learning_rate": 1.4209330530916165e-05, | |
| "loss": 0.0015, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.019413105962450914, | |
| "learning_rate": 1.4082989706067461e-05, | |
| "loss": 0.0008, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.06390145660785629, | |
| "learning_rate": 1.3957170506116201e-05, | |
| "loss": 0.0009, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 0.017555521429492564, | |
| "learning_rate": 1.3831873694944031e-05, | |
| "loss": 0.0007, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 0.04810371136355367, | |
| "learning_rate": 1.3707100033261034e-05, | |
| "loss": 0.0014, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_blimp_filtered_avg": 0.716865671641791, | |
| "eval_blimp_filtered_std": 0.005034673021350593, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_blimp_supplement_avg": 0.834051724137931, | |
| "eval_blimp_supplement_std": 0.016524406240927558, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_vqa_filtered_avg": 0.48, | |
| "eval_vqa_filtered_std": 0.05021167315686779, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_winoground_filtered_avg": 0.68, | |
| "eval_winoground_filtered_std": 0.046882617226215034, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 0.013464323565587763, | |
| "learning_rate": 1.3582850278601134e-05, | |
| "loss": 0.0006, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 0.029333816746967, | |
| "learning_rate": 1.3459125185317434e-05, | |
| "loss": 0.0009, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "grad_norm": 0.019738111761638182, | |
| "learning_rate": 1.3335925504577717e-05, | |
| "loss": 0.0007, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.038558679072453785, | |
| "learning_rate": 1.3213251984359831e-05, | |
| "loss": 0.0009, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.10728075730756652, | |
| "learning_rate": 1.3091105369447165e-05, | |
| "loss": 0.0014, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 0.03939456309548406, | |
| "learning_rate": 1.2969486401424169e-05, | |
| "loss": 0.0009, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 0.02688561690752942, | |
| "learning_rate": 1.2848395818671687e-05, | |
| "loss": 0.0007, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 0.10118668815662003, | |
| "learning_rate": 1.2727834356362778e-05, | |
| "loss": 0.0035, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 0.029540871735792434, | |
| "learning_rate": 1.2607802746457897e-05, | |
| "loss": 0.0008, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 0.019004632066757458, | |
| "learning_rate": 1.2488301717700735e-05, | |
| "loss": 0.0008, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.22, | |
| "grad_norm": 0.050506924718716324, | |
| "learning_rate": 1.2369331995613665e-05, | |
| "loss": 0.0015, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 0.02906757664336032, | |
| "learning_rate": 1.2250894302493265e-05, | |
| "loss": 0.0009, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 0.04633777117624858, | |
| "learning_rate": 1.21329893574062e-05, | |
| "loss": 0.0013, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.03520147966889168, | |
| "learning_rate": 1.2015617876184527e-05, | |
| "loss": 0.0007, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.02696880607305335, | |
| "learning_rate": 1.1898780571421552e-05, | |
| "loss": 0.0009, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.028899261052257607, | |
| "learning_rate": 1.1782478152467502e-05, | |
| "loss": 0.0009, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 0.025787842105068102, | |
| "learning_rate": 1.166671132542505e-05, | |
| "loss": 0.0007, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 0.023827728319133067, | |
| "learning_rate": 1.1551480793145331e-05, | |
| "loss": 0.0009, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 0.030668462671104377, | |
| "learning_rate": 1.1436787255223302e-05, | |
| "loss": 0.0007, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 0.02229439145398785, | |
| "learning_rate": 1.1322631407993811e-05, | |
| "loss": 0.0008, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 0.08225114568135586, | |
| "learning_rate": 1.1209013944527203e-05, | |
| "loss": 0.0017, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.0575147043161383, | |
| "learning_rate": 1.1095935554625148e-05, | |
| "loss": 0.0011, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.05879183520465905, | |
| "learning_rate": 1.098339692481648e-05, | |
| "loss": 0.0016, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.11651577381341721, | |
| "learning_rate": 1.0871398738352955e-05, | |
| "loss": 0.0017, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "grad_norm": 0.04216428635525978, | |
| "learning_rate": 1.0759941675205221e-05, | |
| "loss": 0.0016, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "grad_norm": 0.03712542340164034, | |
| "learning_rate": 1.0649026412058583e-05, | |
| "loss": 0.0012, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 0.04116733983643567, | |
| "learning_rate": 1.0538653622308948e-05, | |
| "loss": 0.0009, | |
| "step": 1127 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 0.05379118870775519, | |
| "learning_rate": 1.042882397605871e-05, | |
| "loss": 0.0014, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 0.023844321094186195, | |
| "learning_rate": 1.0319538140112728e-05, | |
| "loss": 0.0008, | |
| "step": 1129 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.02336448873502654, | |
| "learning_rate": 1.0210796777974197e-05, | |
| "loss": 0.0007, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.04194993258789071, | |
| "learning_rate": 1.0102600549840701e-05, | |
| "loss": 0.0013, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.054557517394749656, | |
| "learning_rate": 9.994950112600154e-06, | |
| "loss": 0.0011, | |
| "step": 1132 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 0.15976956920768468, | |
| "learning_rate": 9.887846119826849e-06, | |
| "loss": 0.0023, | |
| "step": 1133 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 0.05469859681916068, | |
| "learning_rate": 9.781289221777478e-06, | |
| "loss": 0.0011, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 0.0330826221094112, | |
| "learning_rate": 9.675280065387116e-06, | |
| "loss": 0.0012, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 0.037994630222178825, | |
| "learning_rate": 9.569819294265414e-06, | |
| "loss": 0.0007, | |
| "step": 1136 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 0.05446067811636603, | |
| "learning_rate": 9.464907548692614e-06, | |
| "loss": 0.0012, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 0.030162059858075463, | |
| "learning_rate": 9.360545465615667e-06, | |
| "loss": 0.0009, | |
| "step": 1138 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 0.01776810562751649, | |
| "learning_rate": 9.256733678644414e-06, | |
| "loss": 0.0006, | |
| "step": 1139 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "grad_norm": 0.07017734423776259, | |
| "learning_rate": 9.153472818047625e-06, | |
| "loss": 0.0012, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 0.049265529448664684, | |
| "learning_rate": 9.05076351074936e-06, | |
| "loss": 0.0009, | |
| "step": 1141 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 0.01911138033873858, | |
| "learning_rate": 8.948606380324941e-06, | |
| "loss": 0.0008, | |
| "step": 1142 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 0.03207512532598277, | |
| "learning_rate": 8.847002046997354e-06, | |
| "loss": 0.0008, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 0.09498054680592705, | |
| "learning_rate": 8.745951127633411e-06, | |
| "loss": 0.0015, | |
| "step": 1144 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 0.02206935138105853, | |
| "learning_rate": 8.645454235739903e-06, | |
| "loss": 0.0007, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 0.03319965177433449, | |
| "learning_rate": 8.54551198146013e-06, | |
| "loss": 0.0007, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 0.02840733607459401, | |
| "learning_rate": 8.44612497156989e-06, | |
| "loss": 0.0008, | |
| "step": 1147 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 0.018408351544896853, | |
| "learning_rate": 8.347293809474054e-06, | |
| "loss": 0.0006, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 0.08136631297548408, | |
| "learning_rate": 8.249019095202736e-06, | |
| "loss": 0.0025, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 0.03770345328525376, | |
| "learning_rate": 8.151301425407699e-06, | |
| "loss": 0.0011, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 0.021828484215553397, | |
| "learning_rate": 8.054141393358816e-06, | |
| "loss": 0.0008, | |
| "step": 1151 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 0.06858372388136494, | |
| "learning_rate": 7.957539588940299e-06, | |
| "loss": 0.0013, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 0.03540771842662069, | |
| "learning_rate": 7.861496598647278e-06, | |
| "loss": 0.0009, | |
| "step": 1153 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 0.06376796414899714, | |
| "learning_rate": 7.76601300558214e-06, | |
| "loss": 0.0017, | |
| "step": 1154 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 0.08541088279435645, | |
| "learning_rate": 7.671089389451058e-06, | |
| "loss": 0.0012, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.01913967487748475, | |
| "learning_rate": 7.576726326560424e-06, | |
| "loss": 0.0008, | |
| "step": 1156 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.01505834145399495, | |
| "learning_rate": 7.482924389813317e-06, | |
| "loss": 0.0006, | |
| "step": 1157 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.029574633808729885, | |
| "learning_rate": 7.389684148706122e-06, | |
| "loss": 0.0008, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 0.07775324960665371, | |
| "learning_rate": 7.2970061693250154e-06, | |
| "loss": 0.0011, | |
| "step": 1159 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 0.03207503686714158, | |
| "learning_rate": 7.204891014342552e-06, | |
| "loss": 0.0009, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 0.02888329550946116, | |
| "learning_rate": 7.113339243014139e-06, | |
| "loss": 0.0008, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 0.0194570206084492, | |
| "learning_rate": 7.022351411174866e-06, | |
| "loss": 0.0008, | |
| "step": 1162 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 0.14398642426432864, | |
| "learning_rate": 6.931928071235894e-06, | |
| "loss": 0.0032, | |
| "step": 1163 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 0.09922204312397419, | |
| "learning_rate": 6.842069772181236e-06, | |
| "loss": 0.0019, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 0.04657668585468913, | |
| "learning_rate": 6.75277705956443e-06, | |
| "loss": 0.001, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 0.014151142624219092, | |
| "learning_rate": 6.664050475505101e-06, | |
| "loss": 0.0006, | |
| "step": 1166 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 0.041876195216166465, | |
| "learning_rate": 6.575890558685882e-06, | |
| "loss": 0.0013, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 0.02777421677729617, | |
| "learning_rate": 6.48829784434889e-06, | |
| "loss": 0.0008, | |
| "step": 1168 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 0.03827771284490116, | |
| "learning_rate": 6.4012728642926845e-06, | |
| "loss": 0.0015, | |
| "step": 1169 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.1320945724487641, | |
| "learning_rate": 6.314816146868952e-06, | |
| "loss": 0.0021, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.07196675390680835, | |
| "learning_rate": 6.228928216979257e-06, | |
| "loss": 0.0027, | |
| "step": 1171 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 0.06301629094647111, | |
| "learning_rate": 6.143609596072008e-06, | |
| "loss": 0.0008, | |
| "step": 1172 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 0.04690844306331864, | |
| "learning_rate": 6.0588608021390655e-06, | |
| "loss": 0.0012, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 0.02355863678935003, | |
| "learning_rate": 5.97468234971279e-06, | |
| "loss": 0.0007, | |
| "step": 1174 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 0.012557456867681805, | |
| "learning_rate": 5.891074749862857e-06, | |
| "loss": 0.0005, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.47, | |
| "grad_norm": 0.06758959915320514, | |
| "learning_rate": 5.80803851019307e-06, | |
| "loss": 0.0013, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.03280577696402171, | |
| "learning_rate": 5.725574134838474e-06, | |
| "loss": 0.0011, | |
| "step": 1177 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.01838083263589963, | |
| "learning_rate": 5.643682124462057e-06, | |
| "loss": 0.0007, | |
| "step": 1178 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.09736850218115797, | |
| "learning_rate": 5.562362976251901e-06, | |
| "loss": 0.0026, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 0.038368083330140405, | |
| "learning_rate": 5.481617183918053e-06, | |
| "loss": 0.0012, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 0.03241919329467971, | |
| "learning_rate": 5.401445237689573e-06, | |
| "loss": 0.0007, | |
| "step": 1181 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 0.019548513203401407, | |
| "learning_rate": 5.321847624311593e-06, | |
| "loss": 0.0007, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.03343420920569663, | |
| "learning_rate": 5.242824827042237e-06, | |
| "loss": 0.0006, | |
| "step": 1183 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.013860278165749494, | |
| "learning_rate": 5.1643773256498164e-06, | |
| "loss": 0.0004, | |
| "step": 1184 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 0.03250235438008899, | |
| "learning_rate": 5.086505596409885e-06, | |
| "loss": 0.0008, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 0.03812286821993014, | |
| "learning_rate": 5.009210112102292e-06, | |
| "loss": 0.001, | |
| "step": 1186 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 0.04724744059732396, | |
| "learning_rate": 4.932491342008383e-06, | |
| "loss": 0.0008, | |
| "step": 1187 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 0.07149621133781368, | |
| "learning_rate": 4.856349751908107e-06, | |
| "loss": 0.0021, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 0.011104684384602798, | |
| "learning_rate": 4.7807858040771924e-06, | |
| "loss": 0.0005, | |
| "step": 1189 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "grad_norm": 0.02671249567166847, | |
| "learning_rate": 4.705799957284351e-06, | |
| "loss": 0.001, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 0.11429993772721231, | |
| "learning_rate": 4.6313926667885035e-06, | |
| "loss": 0.0019, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 0.019599526292067716, | |
| "learning_rate": 4.557564384335977e-06, | |
| "loss": 0.0007, | |
| "step": 1192 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 0.023058983963623273, | |
| "learning_rate": 4.4843155581578474e-06, | |
| "loss": 0.0007, | |
| "step": 1193 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 0.01789699934131272, | |
| "learning_rate": 4.411646632967059e-06, | |
| "loss": 0.0006, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 0.03475695043132507, | |
| "learning_rate": 4.339558049955927e-06, | |
| "loss": 0.0009, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 0.04626061457798834, | |
| "learning_rate": 4.268050246793276e-06, | |
| "loss": 0.001, | |
| "step": 1196 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 0.03777665253578553, | |
| "learning_rate": 4.197123657621915e-06, | |
| "loss": 0.0009, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 0.08401265052642046, | |
| "learning_rate": 4.126778713055923e-06, | |
| "loss": 0.0018, | |
| "step": 1198 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 0.01795488576011003, | |
| "learning_rate": 4.0570158401780465e-06, | |
| "loss": 0.0006, | |
| "step": 1199 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "grad_norm": 0.05169945359266721, | |
| "learning_rate": 3.987835462537193e-06, | |
| "loss": 0.0012, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "eval_blimp_filtered_avg": 0.7150746268656717, | |
| "eval_blimp_filtered_std": 0.0050445406050592155, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "eval_blimp_supplement_avg": 0.8362068965517241, | |
| "eval_blimp_supplement_std": 0.016487870687225285, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "eval_vqa_filtered_avg": 0.51, | |
| "eval_vqa_filtered_std": 0.05024183937956912, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "eval_winoground_filtered_avg": 0.67, | |
| "eval_winoground_filtered_std": 0.04725815626252606, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 0.03549591031661078, | |
| "learning_rate": 3.919238000145687e-06, | |
| "loss": 0.0014, | |
| "step": 1201 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 0.12212372267764961, | |
| "learning_rate": 3.8512238694768835e-06, | |
| "loss": 0.0018, | |
| "step": 1202 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 0.0708069932236777, | |
| "learning_rate": 3.783793483462583e-06, | |
| "loss": 0.0016, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 0.07637524055458063, | |
| "learning_rate": 3.7169472514904544e-06, | |
| "loss": 0.001, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 0.03487034956626954, | |
| "learning_rate": 3.6506855794016913e-06, | |
| "loss": 0.0012, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 0.026640228769618846, | |
| "learning_rate": 3.5850088694884156e-06, | |
| "loss": 0.0009, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 0.012878283131233357, | |
| "learning_rate": 3.5199175204913117e-06, | |
| "loss": 0.0005, | |
| "step": 1207 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 0.04703267681123701, | |
| "learning_rate": 3.455411927597185e-06, | |
| "loss": 0.0012, | |
| "step": 1208 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.0153560104298258, | |
| "learning_rate": 3.3914924824365422e-06, | |
| "loss": 0.0006, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.05920573168766921, | |
| "learning_rate": 3.3281595730812575e-06, | |
| "loss": 0.001, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.07240573247327886, | |
| "learning_rate": 3.265413584042165e-06, | |
| "loss": 0.0015, | |
| "step": 1211 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 0.03877978634973686, | |
| "learning_rate": 3.203254896266761e-06, | |
| "loss": 0.0009, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 0.0201160866989832, | |
| "learning_rate": 3.1416838871368924e-06, | |
| "loss": 0.0009, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 0.048187931567094715, | |
| "learning_rate": 3.080700930466429e-06, | |
| "loss": 0.0012, | |
| "step": 1214 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 0.07174256245695844, | |
| "learning_rate": 3.0203063964990617e-06, | |
| "loss": 0.0018, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 0.03188905472254505, | |
| "learning_rate": 2.9605006519059507e-06, | |
| "loss": 0.0009, | |
| "step": 1216 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 0.026587740185543742, | |
| "learning_rate": 2.9012840597836045e-06, | |
| "loss": 0.0007, | |
| "step": 1217 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 0.051838257669286805, | |
| "learning_rate": 2.8426569796516146e-06, | |
| "loss": 0.0011, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 0.027884361782616187, | |
| "learning_rate": 2.7846197674504913e-06, | |
| "loss": 0.0007, | |
| "step": 1219 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 0.05934491011655865, | |
| "learning_rate": 2.7271727755395214e-06, | |
| "loss": 0.0014, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 0.07722505492466579, | |
| "learning_rate": 2.6703163526945794e-06, | |
| "loss": 0.0014, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 0.017911785319840934, | |
| "learning_rate": 2.614050844106042e-06, | |
| "loss": 0.0006, | |
| "step": 1222 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 0.07048675908478848, | |
| "learning_rate": 2.5583765913766987e-06, | |
| "loss": 0.0012, | |
| "step": 1223 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 0.029385343497706086, | |
| "learning_rate": 2.503293932519668e-06, | |
| "loss": 0.0011, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 0.026721071049153142, | |
| "learning_rate": 2.4488032019563402e-06, | |
| "loss": 0.0008, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 0.05733009421653368, | |
| "learning_rate": 2.394904730514358e-06, | |
| "loss": 0.0008, | |
| "step": 1226 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 0.048459011121664085, | |
| "learning_rate": 2.3415988454255855e-06, | |
| "loss": 0.001, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 0.08348067589435784, | |
| "learning_rate": 2.2888858703241866e-06, | |
| "loss": 0.0015, | |
| "step": 1228 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 0.0723135438217715, | |
| "learning_rate": 2.236766125244549e-06, | |
| "loss": 0.0013, | |
| "step": 1229 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 0.043987088946181196, | |
| "learning_rate": 2.1852399266194314e-06, | |
| "loss": 0.0012, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 0.02610918331075917, | |
| "learning_rate": 2.1343075872780396e-06, | |
| "loss": 0.0008, | |
| "step": 1231 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 0.02610217806342308, | |
| "learning_rate": 2.0839694164440425e-06, | |
| "loss": 0.0007, | |
| "step": 1232 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 0.016676961400433666, | |
| "learning_rate": 2.03422571973384e-06, | |
| "loss": 0.0007, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 0.0712789068176972, | |
| "learning_rate": 1.985076799154528e-06, | |
| "loss": 0.0011, | |
| "step": 1234 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.040623193014393696, | |
| "learning_rate": 1.9365229531022264e-06, | |
| "loss": 0.0009, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.0174120811798264, | |
| "learning_rate": 1.8885644763601774e-06, | |
| "loss": 0.0006, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.02397234113764592, | |
| "learning_rate": 1.8412016600969695e-06, | |
| "loss": 0.0008, | |
| "step": 1237 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 0.04262503636577534, | |
| "learning_rate": 1.7944347918648185e-06, | |
| "loss": 0.0011, | |
| "step": 1238 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 0.02559933629938596, | |
| "learning_rate": 1.748264155597712e-06, | |
| "loss": 0.0007, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 0.024338642632736805, | |
| "learning_rate": 1.7026900316098215e-06, | |
| "loss": 0.0007, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 0.04865116104780992, | |
| "learning_rate": 1.657712696593705e-06, | |
| "loss": 0.0014, | |
| "step": 1241 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 0.016653138748607372, | |
| "learning_rate": 1.6133324236186742e-06, | |
| "loss": 0.0006, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.02913263827559182, | |
| "learning_rate": 1.5695494821290735e-06, | |
| "loss": 0.0007, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.1053160720503692, | |
| "learning_rate": 1.5263641379427595e-06, | |
| "loss": 0.0022, | |
| "step": 1244 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.06917659494864752, | |
| "learning_rate": 1.4837766532493468e-06, | |
| "loss": 0.0014, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 0.034919075850131744, | |
| "learning_rate": 1.4417872866087534e-06, | |
| "loss": 0.0006, | |
| "step": 1246 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 0.07398471019830524, | |
| "learning_rate": 1.400396292949513e-06, | |
| "loss": 0.0013, | |
| "step": 1247 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.02043767100243107, | |
| "learning_rate": 1.3596039235672874e-06, | |
| "loss": 0.0007, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.026716481737152874, | |
| "learning_rate": 1.3194104261233798e-06, | |
| "loss": 0.0007, | |
| "step": 1249 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.04718046174374857, | |
| "learning_rate": 1.2798160446431006e-06, | |
| "loss": 0.0013, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 0.053929792186028926, | |
| "learning_rate": 1.2408210195144376e-06, | |
| "loss": 0.001, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 0.016211755152668282, | |
| "learning_rate": 1.2024255874865108e-06, | |
| "loss": 0.0006, | |
| "step": 1252 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "grad_norm": 0.02218925262348946, | |
| "learning_rate": 1.1646299816681195e-06, | |
| "loss": 0.0006, | |
| "step": 1253 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 0.061559821718042196, | |
| "learning_rate": 1.1274344315264196e-06, | |
| "loss": 0.0018, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 0.011397207501113183, | |
| "learning_rate": 1.0908391628854041e-06, | |
| "loss": 0.0005, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 0.022367382815004444, | |
| "learning_rate": 1.0548443979246481e-06, | |
| "loss": 0.0008, | |
| "step": 1256 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 0.11917770316499016, | |
| "learning_rate": 1.0194503551778866e-06, | |
| "loss": 0.0016, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "grad_norm": 0.023235418948401518, | |
| "learning_rate": 9.846572495316952e-07, | |
| "loss": 0.0007, | |
| "step": 1258 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 0.017722263245702797, | |
| "learning_rate": 9.504652922242562e-07, | |
| "loss": 0.0007, | |
| "step": 1259 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 0.15628375169850256, | |
| "learning_rate": 9.168746908439718e-07, | |
| "loss": 0.0026, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 0.019997091181886018, | |
| "learning_rate": 8.838856493282754e-07, | |
| "loss": 0.0007, | |
| "step": 1261 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.05125265984626611, | |
| "learning_rate": 8.514983679623556e-07, | |
| "loss": 0.001, | |
| "step": 1262 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.014670438570475425, | |
| "learning_rate": 8.197130433779565e-07, | |
| "loss": 0.0005, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 0.06613452163795065, | |
| "learning_rate": 7.885298685522235e-07, | |
| "loss": 0.0024, | |
| "step": 1264 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 0.03851957779319382, | |
| "learning_rate": 7.579490328064265e-07, | |
| "loss": 0.0013, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 0.01550807081268757, | |
| "learning_rate": 7.27970721804927e-07, | |
| "loss": 0.0006, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 0.02645554854912103, | |
| "learning_rate": 6.985951175539796e-07, | |
| "loss": 0.0007, | |
| "step": 1267 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 0.03279121203073833, | |
| "learning_rate": 6.698223984006436e-07, | |
| "loss": 0.0009, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 0.0519224271730649, | |
| "learning_rate": 6.416527390317173e-07, | |
| "loss": 0.0009, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 0.09986357170391567, | |
| "learning_rate": 6.140863104726391e-07, | |
| "loss": 0.0016, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 0.03275074991380398, | |
| "learning_rate": 5.87123280086499e-07, | |
| "loss": 0.0007, | |
| "step": 1271 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 0.04587670912561613, | |
| "learning_rate": 5.607638115729841e-07, | |
| "loss": 0.0015, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 0.027686155223001784, | |
| "learning_rate": 5.350080649674127e-07, | |
| "loss": 0.0007, | |
| "step": 1273 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "grad_norm": 0.0223211970919243, | |
| "learning_rate": 5.098561966397131e-07, | |
| "loss": 0.0006, | |
| "step": 1274 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 0.017928512348313316, | |
| "learning_rate": 4.85308359293557e-07, | |
| "loss": 0.0005, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 0.03475070764217293, | |
| "learning_rate": 4.613647019653389e-07, | |
| "loss": 0.0007, | |
| "step": 1276 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 0.020434320555203736, | |
| "learning_rate": 4.3802537002335386e-07, | |
| "loss": 0.0008, | |
| "step": 1277 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 0.025932942114901558, | |
| "learning_rate": 4.152905051668321e-07, | |
| "loss": 0.0008, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 0.0629734875825493, | |
| "learning_rate": 3.931602454251837e-07, | |
| "loss": 0.0014, | |
| "step": 1279 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 0.017477717878726103, | |
| "learning_rate": 3.716347251570551e-07, | |
| "loss": 0.0006, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 0.028129693282624404, | |
| "learning_rate": 3.50714075049563e-07, | |
| "loss": 0.0009, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 0.3789011247756691, | |
| "learning_rate": 3.3039842211752824e-07, | |
| "loss": 0.0007, | |
| "step": 1282 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 0.02028981654233484, | |
| "learning_rate": 3.106878897026544e-07, | |
| "loss": 0.0007, | |
| "step": 1283 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 0.017043915032762993, | |
| "learning_rate": 2.915825974727726e-07, | |
| "loss": 0.0007, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 0.02765997006700044, | |
| "learning_rate": 2.7308266142119785e-07, | |
| "loss": 0.0008, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 0.05844451311284074, | |
| "learning_rate": 2.5518819386590734e-07, | |
| "loss": 0.0011, | |
| "step": 1286 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 0.05688879982444352, | |
| "learning_rate": 2.3789930344897404e-07, | |
| "loss": 0.0008, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.028207898085014158, | |
| "learning_rate": 2.212160951358011e-07, | |
| "loss": 0.001, | |
| "step": 1288 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.016086514604394798, | |
| "learning_rate": 2.0513867021457744e-07, | |
| "loss": 0.0007, | |
| "step": 1289 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.02971312168354947, | |
| "learning_rate": 1.8966712629558957e-07, | |
| "loss": 0.0009, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 0.021944051433958625, | |
| "learning_rate": 1.748015573106887e-07, | |
| "loss": 0.0007, | |
| "step": 1291 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 0.04949653596283253, | |
| "learning_rate": 1.6054205351265784e-07, | |
| "loss": 0.0021, | |
| "step": 1292 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 0.06054720833614147, | |
| "learning_rate": 1.4688870147473443e-07, | |
| "loss": 0.0011, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 0.02434476920884362, | |
| "learning_rate": 1.338415840900109e-07, | |
| "loss": 0.0007, | |
| "step": 1294 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "grad_norm": 0.04425200814298354, | |
| "learning_rate": 1.2140078057101266e-07, | |
| "loss": 0.001, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 0.08400768778349636, | |
| "learning_rate": 1.0956636644912088e-07, | |
| "loss": 0.0015, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "grad_norm": 0.01681668533811899, | |
| "learning_rate": 9.833841357421714e-08, | |
| "loss": 0.0006, | |
| "step": 1297 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 0.06053068555064212, | |
| "learning_rate": 8.771699011416168e-08, | |
| "loss": 0.0019, | |
| "step": 1298 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 0.04317850317270507, | |
| "learning_rate": 7.770216055443814e-08, | |
| "loss": 0.0011, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 0.08586897126139875, | |
| "learning_rate": 6.829398569770939e-08, | |
| "loss": 0.001, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_blimp_filtered_avg": 0.7153731343283583, | |
| "eval_blimp_filtered_std": 0.005045209651150972, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_blimp_supplement_avg": 0.834051724137931, | |
| "eval_blimp_supplement_std": 0.016654601081150626, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_vqa_filtered_avg": 0.5, | |
| "eval_vqa_filtered_std": 0.050251890762960605, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_winoground_filtered_avg": 0.69, | |
| "eval_winoground_filtered_std": 0.04648231987117316, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 0.016208779457975556, | |
| "learning_rate": 5.94925226635068e-08, | |
| "loss": 0.0006, | |
| "step": 1301 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 0.01833907339793384, | |
| "learning_rate": 5.1297824887841516e-08, | |
| "loss": 0.0006, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 0.036830940713068985, | |
| "learning_rate": 4.370994212287149e-08, | |
| "loss": 0.001, | |
| "step": 1303 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 0.015645769111424807, | |
| "learning_rate": 3.672892043666831e-08, | |
| "loss": 0.0006, | |
| "step": 1304 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 0.07940082249502824, | |
| "learning_rate": 3.0354802212839705e-08, | |
| "loss": 0.0011, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 0.015159814855124942, | |
| "learning_rate": 2.4587626150351926e-08, | |
| "loss": 0.0005, | |
| "step": 1306 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 0.022458919214575546, | |
| "learning_rate": 1.94274272632633e-08, | |
| "loss": 0.001, | |
| "step": 1307 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 0.02980953425255622, | |
| "learning_rate": 1.4874236880491055e-08, | |
| "loss": 0.0008, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 0.023911967525172478, | |
| "learning_rate": 1.0928082645667025e-08, | |
| "loss": 0.0007, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 0.01333819056514401, | |
| "learning_rate": 7.58898851693779e-09, | |
| "loss": 0.0005, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 0.028702718224002614, | |
| "learning_rate": 4.856974766831446e-09, | |
| "loss": 0.0008, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 0.022211928652402655, | |
| "learning_rate": 2.732057982124392e-09, | |
| "loss": 0.0006, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "grad_norm": 0.021399146337567086, | |
| "learning_rate": 1.2142510637414006e-09, | |
| "loss": 0.0008, | |
| "step": 1313 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.06401289450470597, | |
| "learning_rate": 3.035632266890076e-10, | |
| "loss": 0.0017, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.10829033858276516, | |
| "learning_rate": 0.0, | |
| "loss": 0.0017, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1315, | |
| "total_flos": 76761234800640.0, | |
| "train_loss": 0.24336853736315273, | |
| "train_runtime": 20724.9349, | |
| "train_samples_per_second": 16.232, | |
| "train_steps_per_second": 0.063 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 1315, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 76761234800640.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |