| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 210, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 80.00438704007676, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 6.2941, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 96.34802446721417, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 6.8085, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 154.2232661357414, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 6.1192, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 23.03476938439163, | |
| "learning_rate": 1.1428571428571429e-05, | |
| "loss": 5.4161, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 11.647036257139195, | |
| "learning_rate": 1.4285714285714287e-05, | |
| "loss": 5.2596, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 14.698112775361619, | |
| "learning_rate": 1.7142857142857142e-05, | |
| "loss": 5.3002, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 23.71107587456011, | |
| "learning_rate": 2e-05, | |
| "loss": 4.6473, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 14.120600306959602, | |
| "learning_rate": 1.9998802517966852e-05, | |
| "loss": 5.3174, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 8.191706198497702, | |
| "learning_rate": 1.9995210358660037e-05, | |
| "loss": 4.9716, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 10.458311244778137, | |
| "learning_rate": 1.9989224382388813e-05, | |
| "loss": 4.5338, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 9.213287452666169, | |
| "learning_rate": 1.9980846022772978e-05, | |
| "loss": 4.7343, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 5.9009430281116675, | |
| "learning_rate": 1.997007728639956e-05, | |
| "loss": 5.0022, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 7.8868800818479174, | |
| "learning_rate": 1.9956920752342226e-05, | |
| "loss": 5.0521, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.259759634256785, | |
| "learning_rate": 1.9941379571543597e-05, | |
| "loss": 4.6051, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 5.579574504065297, | |
| "learning_rate": 1.9923457466060637e-05, | |
| "loss": 4.2182, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 5.913831283576424, | |
| "learning_rate": 1.9903158728173206e-05, | |
| "loss": 4.2917, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 7.717836608401082, | |
| "learning_rate": 1.9880488219356086e-05, | |
| "loss": 4.3894, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 6.296015368993091, | |
| "learning_rate": 1.9855451369114677e-05, | |
| "loss": 4.3322, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 8.475421651603199, | |
| "learning_rate": 1.9828054173684646e-05, | |
| "loss": 4.6874, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 8.452753919149812, | |
| "learning_rate": 1.9798303194595846e-05, | |
| "loss": 4.3446, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 5.716694629623973, | |
| "learning_rate": 1.976620555710087e-05, | |
| "loss": 4.0236, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 6.3915589352407345, | |
| "learning_rate": 1.973176894846855e-05, | |
| "loss": 4.4564, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 3.849187023823461, | |
| "learning_rate": 1.9695001616142916e-05, | |
| "loss": 3.9989, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.228918174540112, | |
| "learning_rate": 1.965591236576794e-05, | |
| "loss": 4.3524, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 5.974882597027604, | |
| "learning_rate": 1.9614510559078626e-05, | |
| "loss": 4.1009, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 4.926401136001884, | |
| "learning_rate": 1.95708061116589e-05, | |
| "loss": 4.216, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.034835120114335, | |
| "learning_rate": 1.9524809490566878e-05, | |
| "loss": 4.3396, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.643514172374943, | |
| "learning_rate": 1.9476531711828027e-05, | |
| "loss": 4.0161, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 4.210788851218648, | |
| "learning_rate": 1.942598433779687e-05, | |
| "loss": 4.5483, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.684561962496697, | |
| "learning_rate": 1.9373179474387858e-05, | |
| "loss": 4.3755, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 4.673387982745991, | |
| "learning_rate": 1.9318129768176033e-05, | |
| "loss": 4.4423, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.990722772080045, | |
| "learning_rate": 1.926084840336821e-05, | |
| "loss": 4.3072, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 4.9483245517186605, | |
| "learning_rate": 1.9201349098645433e-05, | |
| "loss": 4.2863, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 4.427315167497816, | |
| "learning_rate": 1.9139646103877378e-05, | |
| "loss": 4.1236, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.922455864121031, | |
| "learning_rate": 1.9075754196709574e-05, | |
| "loss": 4.7598, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 4.123279198243713, | |
| "learning_rate": 1.900968867902419e-05, | |
| "loss": 3.8454, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 4.251851688874165, | |
| "learning_rate": 1.894146537327533e-05, | |
| "loss": 4.1765, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 4.440763895011154, | |
| "learning_rate": 1.8871100618699553e-05, | |
| "loss": 4.5498, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 4.359765853893123, | |
| "learning_rate": 1.8798611267402745e-05, | |
| "loss": 4.5428, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 5.0972668892147475, | |
| "learning_rate": 1.872401468032406e-05, | |
| "loss": 4.5734, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 5.452699114764846, | |
| "learning_rate": 1.864732872307804e-05, | |
| "loss": 4.083, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.6602209793027534, | |
| "learning_rate": 1.8568571761675893e-05, | |
| "loss": 3.8202, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 4.930178982421816, | |
| "learning_rate": 1.8487762658126872e-05, | |
| "loss": 4.2528, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 4.407664636029237, | |
| "learning_rate": 1.8404920765920898e-05, | |
| "loss": 3.8699, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 4.78545005474896, | |
| "learning_rate": 1.8320065925393468e-05, | |
| "loss": 3.6822, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 4.078401960467532, | |
| "learning_rate": 1.8233218458973984e-05, | |
| "loss": 3.7794, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 5.326820494758211, | |
| "learning_rate": 1.814439916631857e-05, | |
| "loss": 3.7876, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 6.378382372761396, | |
| "learning_rate": 1.8053629319328662e-05, | |
| "loss": 3.9788, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 5.750592810925643, | |
| "learning_rate": 1.796093065705644e-05, | |
| "loss": 3.6474, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 3.8174994985593544, | |
| "learning_rate": 1.786632538049842e-05, | |
| "loss": 3.3771, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 5.557487394414142, | |
| "learning_rate": 1.776983614727838e-05, | |
| "loss": 3.3717, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 4.074429527562, | |
| "learning_rate": 1.7671486066220965e-05, | |
| "loss": 3.5966, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 4.063021464956913, | |
| "learning_rate": 1.757129869181718e-05, | |
| "loss": 3.507, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 4.894442593636845, | |
| "learning_rate": 1.746929801858317e-05, | |
| "loss": 3.9196, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 4.452722126818571, | |
| "learning_rate": 1.736550847531366e-05, | |
| "loss": 3.6257, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 4.00522098752344, | |
| "learning_rate": 1.725995491923131e-05, | |
| "loss": 3.3246, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 4.25230998576677, | |
| "learning_rate": 1.7152662630033506e-05, | |
| "loss": 3.6398, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 4.089951263233628, | |
| "learning_rate": 1.7043657303837965e-05, | |
| "loss": 3.1908, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 5.72352434378825, | |
| "learning_rate": 1.693296504702862e-05, | |
| "loss": 3.5252, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 5.160209312433058, | |
| "learning_rate": 1.682061237000322e-05, | |
| "loss": 3.4718, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 5.488539699634776, | |
| "learning_rate": 1.6706626180824185e-05, | |
| "loss": 3.4474, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 4.42082715558397, | |
| "learning_rate": 1.659103377877423e-05, | |
| "loss": 3.4319, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 5.021528964342142, | |
| "learning_rate": 1.647386284781828e-05, | |
| "loss": 3.2333, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 4.771790748100547, | |
| "learning_rate": 1.6355141449973254e-05, | |
| "loss": 3.5741, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 4.315875791980731, | |
| "learning_rate": 1.6234898018587336e-05, | |
| "loss": 3.565, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 4.243056541134396, | |
| "learning_rate": 1.611316135153026e-05, | |
| "loss": 3.5365, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 4.346968517779339, | |
| "learning_rate": 1.598996060429634e-05, | |
| "loss": 3.5757, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 4.884427461945468, | |
| "learning_rate": 1.586532528302183e-05, | |
| "loss": 3.5492, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 4.5157468490886785, | |
| "learning_rate": 1.5739285237418323e-05, | |
| "loss": 3.3304, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.070256806343138, | |
| "learning_rate": 1.5611870653623826e-05, | |
| "loss": 3.9314, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 3.6449860141474963, | |
| "learning_rate": 1.548311204697331e-05, | |
| "loss": 3.1916, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 4.255978041595801, | |
| "learning_rate": 1.5353040254690396e-05, | |
| "loss": 3.4329, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 5.139952618469592, | |
| "learning_rate": 1.5221686428501929e-05, | |
| "loss": 3.6179, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 4.605357638561885, | |
| "learning_rate": 1.5089082027177291e-05, | |
| "loss": 3.6597, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 6.077382699948124, | |
| "learning_rate": 1.4955258808994096e-05, | |
| "loss": 3.6931, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 5.497332973780629, | |
| "learning_rate": 1.4820248824132221e-05, | |
| "loss": 4.3554, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 5.805455693303122, | |
| "learning_rate": 1.4684084406997903e-05, | |
| "loss": 4.3734, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 5.457678962357698, | |
| "learning_rate": 1.4546798168479756e-05, | |
| "loss": 4.4448, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 4.520222765843399, | |
| "learning_rate": 1.4408422988138585e-05, | |
| "loss": 4.347, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 5.401119547471518, | |
| "learning_rate": 1.4268992006332847e-05, | |
| "loss": 4.5233, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 6.4460955298094955, | |
| "learning_rate": 1.412853861628166e-05, | |
| "loss": 3.1766, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 5.789832740451367, | |
| "learning_rate": 1.3987096456067236e-05, | |
| "loss": 3.3281, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 4.998279839354735, | |
| "learning_rate": 1.3844699400578696e-05, | |
| "loss": 3.378, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 4.615720057294197, | |
| "learning_rate": 1.3701381553399147e-05, | |
| "loss": 3.0408, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 4.469304200887987, | |
| "learning_rate": 1.3557177238637987e-05, | |
| "loss": 2.7044, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 4.222990961364449, | |
| "learning_rate": 1.3412120992710425e-05, | |
| "loss": 2.8298, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 4.805325799209531, | |
| "learning_rate": 1.3266247556066122e-05, | |
| "loss": 2.7601, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 5.480294735880267, | |
| "learning_rate": 1.3119591864868979e-05, | |
| "loss": 2.7043, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 5.768438764027398, | |
| "learning_rate": 1.2972189042630044e-05, | |
| "loss": 3.0128, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 6.141508752253715, | |
| "learning_rate": 1.2824074391795571e-05, | |
| "loss": 2.7125, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 5.6316708738751435, | |
| "learning_rate": 1.2675283385292212e-05, | |
| "loss": 2.6139, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 5.679660349214387, | |
| "learning_rate": 1.252585165803135e-05, | |
| "loss": 2.7675, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 5.145903533309325, | |
| "learning_rate": 1.2375814998374714e-05, | |
| "loss": 2.6252, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 5.421011319827069, | |
| "learning_rate": 1.2225209339563144e-05, | |
| "loss": 2.8014, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 5.468955060971493, | |
| "learning_rate": 1.2074070751110752e-05, | |
| "loss": 2.837, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 5.164937111950548, | |
| "learning_rate": 1.192243543016637e-05, | |
| "loss": 2.8372, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 5.180488985174343, | |
| "learning_rate": 1.1770339692844484e-05, | |
| "loss": 2.547, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 6.006405690750611, | |
| "learning_rate": 1.161781996552765e-05, | |
| "loss": 2.5804, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 5.952437046746775, | |
| "learning_rate": 1.1464912776142494e-05, | |
| "loss": 2.7432, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 5.094786893941399, | |
| "learning_rate": 1.1311654745411424e-05, | |
| "loss": 2.6715, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 5.351854924535129, | |
| "learning_rate": 1.115808257808209e-05, | |
| "loss": 2.5912, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 5.278029090648024, | |
| "learning_rate": 1.1004233054136726e-05, | |
| "loss": 2.7294, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 5.602710570648003, | |
| "learning_rate": 1.0850143019983475e-05, | |
| "loss": 2.8212, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 5.203710653440999, | |
| "learning_rate": 1.0695849379631813e-05, | |
| "loss": 2.6128, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 5.712401092804933, | |
| "learning_rate": 1.0541389085854177e-05, | |
| "loss": 3.1367, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 4.625678226976654, | |
| "learning_rate": 1.038679913133589e-05, | |
| "loss": 2.4625, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 4.945017754425003, | |
| "learning_rate": 1.0232116539815558e-05, | |
| "loss": 2.6775, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 5.317308116977884, | |
| "learning_rate": 1.0077378357218022e-05, | |
| "loss": 2.8065, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 5.417094774783667, | |
| "learning_rate": 9.92262164278198e-06, | |
| "loss": 2.8207, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 5.470263884411447, | |
| "learning_rate": 9.767883460184444e-06, | |
| "loss": 2.9305, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 5.098945198369544, | |
| "learning_rate": 9.613200868664112e-06, | |
| "loss": 3.4538, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 6.026048110076312, | |
| "learning_rate": 9.458610914145826e-06, | |
| "loss": 3.506, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 5.820084192866225, | |
| "learning_rate": 9.304150620368189e-06, | |
| "loss": 3.847, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "grad_norm": 5.454079563635766, | |
| "learning_rate": 9.149856980016529e-06, | |
| "loss": 3.6699, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 4.856647591410433, | |
| "learning_rate": 8.995766945863278e-06, | |
| "loss": 3.9941, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 5.445692392998836, | |
| "learning_rate": 8.841917421917913e-06, | |
| "loss": 4.3554, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 5.084555008534811, | |
| "learning_rate": 8.688345254588579e-06, | |
| "loss": 4.3838, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 4.985430311007982, | |
| "learning_rate": 8.53508722385751e-06, | |
| "loss": 4.5445, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 3.943933363518215, | |
| "learning_rate": 8.382180034472353e-06, | |
| "loss": 4.2051, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 4.491172285358988, | |
| "learning_rate": 8.229660307155518e-06, | |
| "loss": 4.498, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 6.770327191921559, | |
| "learning_rate": 8.077564569833633e-06, | |
| "loss": 2.5859, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 6.344093649426968, | |
| "learning_rate": 7.92592924888925e-06, | |
| "loss": 2.6089, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 5.542543063619591, | |
| "learning_rate": 7.774790660436857e-06, | |
| "loss": 2.5915, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 5.620129695091442, | |
| "learning_rate": 7.6241850016252915e-06, | |
| "loss": 2.7973, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 4.937221142463724, | |
| "learning_rate": 7.4741483419686525e-06, | |
| "loss": 2.4647, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 4.177317988292998, | |
| "learning_rate": 7.324716614707794e-06, | |
| "loss": 2.3549, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 4.873379259329313, | |
| "learning_rate": 7.175925608204428e-06, | |
| "loss": 2.3888, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.66, | |
| "grad_norm": 4.027921720790141, | |
| "learning_rate": 7.0278109573699574e-06, | |
| "loss": 2.2492, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 4.489654053696505, | |
| "learning_rate": 6.880408135131022e-06, | |
| "loss": 2.244, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 4.515410453085113, | |
| "learning_rate": 6.733752443933879e-06, | |
| "loss": 2.3124, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 4.963204172080167, | |
| "learning_rate": 6.587879007289576e-06, | |
| "loss": 2.2542, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "grad_norm": 4.931378399728066, | |
| "learning_rate": 6.442822761362015e-06, | |
| "loss": 1.9509, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 4.7328764276001, | |
| "learning_rate": 6.298618446600856e-06, | |
| "loss": 2.0579, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 5.251697085848756, | |
| "learning_rate": 6.155300599421305e-06, | |
| "loss": 2.1462, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 3.86, | |
| "grad_norm": 5.377799005102049, | |
| "learning_rate": 6.0129035439327665e-06, | |
| "loss": 1.991, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 5.185963077105245, | |
| "learning_rate": 5.871461383718344e-06, | |
| "loss": 1.8895, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 5.590606159281845, | |
| "learning_rate": 5.731007993667155e-06, | |
| "loss": 2.0092, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 6.164057695091001, | |
| "learning_rate": 5.5915770118614195e-06, | |
| "loss": 2.0822, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 6.77180320115024, | |
| "learning_rate": 5.453201831520245e-06, | |
| "loss": 1.8797, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.052961110479867, | |
| "learning_rate": 5.3159155930021e-06, | |
| "loss": 2.3115, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 13.172067376591047, | |
| "learning_rate": 5.179751175867784e-06, | |
| "loss": 5.3575, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 13.334878110202311, | |
| "learning_rate": 5.044741191005908e-06, | |
| "loss": 5.6717, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 10.171581292048922, | |
| "learning_rate": 4.910917972822713e-06, | |
| "loss": 4.9323, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 8.877047299036951, | |
| "learning_rate": 4.778313571498075e-06, | |
| "loss": 4.5247, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 7.916260375247708, | |
| "learning_rate": 4.64695974530961e-06, | |
| "loss": 4.2104, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 7.001700234460781, | |
| "learning_rate": 4.5168879530266906e-06, | |
| "loss": 3.6699, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 5.344925398470186, | |
| "learning_rate": 4.388129346376177e-06, | |
| "loss": 3.3075, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 5.740352187809239, | |
| "learning_rate": 4.260714762581677e-06, | |
| "loss": 3.5958, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 4.791723313221749, | |
| "learning_rate": 4.1346747169781695e-06, | |
| "loss": 3.2025, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 4.6059266231606655, | |
| "learning_rate": 4.010039395703665e-06, | |
| "loss": 3.0045, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 5.208392250257202, | |
| "learning_rate": 3.886838648469742e-06, | |
| "loss": 2.5741, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 5.233522675196091, | |
| "learning_rate": 3.7651019814126656e-06, | |
| "loss": 2.6712, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 5.135869425784653, | |
| "learning_rate": 3.6448585500267487e-06, | |
| "loss": 2.6455, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 4.645208008447359, | |
| "learning_rate": 3.5261371521817247e-06, | |
| "loss": 2.3787, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "grad_norm": 4.660492061746799, | |
| "learning_rate": 3.4089662212257733e-06, | |
| "loss": 2.1385, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 4.4175550152417635, | |
| "learning_rate": 3.2933738191758158e-06, | |
| "loss": 2.0405, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "grad_norm": 4.566709555922203, | |
| "learning_rate": 3.179387629996782e-06, | |
| "loss": 1.9903, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 4.349261816616526, | |
| "learning_rate": 3.067034952971382e-06, | |
| "loss": 1.9797, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "grad_norm": 4.815462187163981, | |
| "learning_rate": 2.956342696162037e-06, | |
| "loss": 2.0851, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "grad_norm": 4.791928073499566, | |
| "learning_rate": 2.8473373699664995e-06, | |
| "loss": 1.793, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 4.320192640779764, | |
| "learning_rate": 2.740045080768694e-06, | |
| "loss": 1.746, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.63, | |
| "grad_norm": 4.604601663113942, | |
| "learning_rate": 2.634491524686341e-06, | |
| "loss": 1.6877, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "grad_norm": 3.63642030530214, | |
| "learning_rate": 2.5307019814168343e-06, | |
| "loss": 1.6653, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 3.7605261098294744, | |
| "learning_rate": 2.428701308182826e-06, | |
| "loss": 1.5892, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 3.770794461990271, | |
| "learning_rate": 2.3285139337790343e-06, | |
| "loss": 1.7293, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 3.9388567675880406, | |
| "learning_rate": 2.2301638527216196e-06, | |
| "loss": 1.6751, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 3.59303494013787, | |
| "learning_rate": 2.1336746195015845e-06, | |
| "loss": 1.3555, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 3.616144561640801, | |
| "learning_rate": 2.0390693429435626e-06, | |
| "loss": 1.4791, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 4.83, | |
| "grad_norm": 4.184463124765985, | |
| "learning_rate": 1.946370680671341e-06, | |
| "loss": 1.5188, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 4.105357361825141, | |
| "learning_rate": 1.8556008336814302e-06, | |
| "loss": 1.4057, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 4.231632132658564, | |
| "learning_rate": 1.7667815410260181e-06, | |
| "loss": 1.3032, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 4.194933198894844, | |
| "learning_rate": 1.679934074606533e-06, | |
| "loss": 1.4148, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 4.675308614956284, | |
| "learning_rate": 1.5950792340791044e-06, | |
| "loss": 1.5084, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 4.92049256107502, | |
| "learning_rate": 1.5122373418731306e-06, | |
| "loss": 1.3565, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.969442514645622, | |
| "learning_rate": 1.4314282383241097e-06, | |
| "loss": 1.7056, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 6.797850651633187, | |
| "learning_rate": 1.3526712769219619e-06, | |
| "loss": 1.9073, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "grad_norm": 7.510748075175952, | |
| "learning_rate": 1.2759853196759454e-06, | |
| "loss": 2.0683, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "grad_norm": 7.689466727133375, | |
| "learning_rate": 1.201388732597255e-06, | |
| "loss": 2.093, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.11, | |
| "grad_norm": 7.170048914177173, | |
| "learning_rate": 1.1288993813004467e-06, | |
| "loss": 2.1694, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.14, | |
| "grad_norm": 6.888698821967011, | |
| "learning_rate": 1.0585346267246743e-06, | |
| "loss": 2.1485, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 8.935327932773875, | |
| "learning_rate": 9.903113209758098e-07, | |
| "loss": 3.2942, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 7.282062148382787, | |
| "learning_rate": 9.242458032904311e-07, | |
| "loss": 3.0451, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.23, | |
| "grad_norm": 8.856238734952932, | |
| "learning_rate": 8.603538961226233e-07, | |
| "loss": 3.2713, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "grad_norm": 7.072514973353356, | |
| "learning_rate": 7.986509013545673e-07, | |
| "loss": 2.8979, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 6.300112859192955, | |
| "learning_rate": 7.391515966317908e-07, | |
| "loss": 2.7407, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 6.552708837647418, | |
| "learning_rate": 6.81870231823969e-07, | |
| "loss": 2.3151, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "grad_norm": 6.390611058615414, | |
| "learning_rate": 6.268205256121396e-07, | |
| "loss": 2.3887, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "grad_norm": 6.1849954718581195, | |
| "learning_rate": 5.740156622031301e-07, | |
| "loss": 2.3422, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 5.499570423135919, | |
| "learning_rate": 5.234682881719766e-07, | |
| "loss": 2.1039, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 4.872882858791267, | |
| "learning_rate": 4.7519050943312326e-07, | |
| "loss": 1.9011, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "grad_norm": 4.373003479063446, | |
| "learning_rate": 4.2919388834110066e-07, | |
| "loss": 1.7516, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "grad_norm": 4.508004752636755, | |
| "learning_rate": 3.854894409213761e-07, | |
| "loss": 1.6404, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.51, | |
| "grad_norm": 4.2889965352379535, | |
| "learning_rate": 3.4408763423206093e-07, | |
| "loss": 1.6365, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "grad_norm": 4.682234330521379, | |
| "learning_rate": 3.049983838570858e-07, | |
| "loss": 1.6745, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 4.537341291350183, | |
| "learning_rate": 2.6823105153145125e-07, | |
| "loss": 1.4181, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 4.32840882506885, | |
| "learning_rate": 2.3379444289913344e-07, | |
| "loss": 1.41, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 5.63, | |
| "grad_norm": 4.618590743310313, | |
| "learning_rate": 2.016968054041546e-07, | |
| "loss": 1.3301, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "grad_norm": 3.9839592670692716, | |
| "learning_rate": 1.7194582631535616e-07, | |
| "loss": 1.3857, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "grad_norm": 4.189042770104091, | |
| "learning_rate": 1.4454863088532388e-07, | |
| "loss": 1.3024, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "grad_norm": 4.2928472832951305, | |
| "learning_rate": 1.19511780643915e-07, | |
| "loss": 1.4741, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "grad_norm": 4.58876530743957, | |
| "learning_rate": 9.684127182679526e-08, | |
| "loss": 1.4256, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "grad_norm": 4.562646086161964, | |
| "learning_rate": 7.65425339393644e-08, | |
| "loss": 1.128, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 4.530949504721316, | |
| "learning_rate": 5.862042845640403e-08, | |
| "loss": 1.2618, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 5.3286756918257865, | |
| "learning_rate": 4.307924765777682e-08, | |
| "loss": 1.2957, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "grad_norm": 5.2853418793857845, | |
| "learning_rate": 2.992271360043986e-08, | |
| "loss": 1.2061, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "grad_norm": 5.532437877616826, | |
| "learning_rate": 1.9153977227022168e-08, | |
| "loss": 1.1273, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "grad_norm": 5.896991609572843, | |
| "learning_rate": 1.0775617611189504e-08, | |
| "loss": 1.2515, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 6.221140999378529, | |
| "learning_rate": 4.7896413399639575e-09, | |
| "loss": 1.348, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 5.97, | |
| "grad_norm": 5.940323080238109, | |
| "learning_rate": 1.1974820331517313e-09, | |
| "loss": 1.2091, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 6.754644623774243, | |
| "learning_rate": 0.0, | |
| "loss": 1.5446, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 210, | |
| "total_flos": 0.0, | |
| "train_loss": 0.06094220252264114, | |
| "train_runtime": 5786.5332, | |
| "train_samples_per_second": 74.324, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 210, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 20, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |