| { | |
| "best_metric": 0.4741531312465668, | |
| "best_model_checkpoint": "NHS-BiomedNLP-BiomedBERT-hypop\\checkpoint-397", | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 397, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 9.947269439697266, | |
| "learning_rate": 2.9987405541561713e-05, | |
| "loss": 0.7027, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 8.814902305603027, | |
| "learning_rate": 2.997481108312343e-05, | |
| "loss": 0.6097, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 14.668534278869629, | |
| "learning_rate": 2.9962216624685138e-05, | |
| "loss": 0.7259, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 6.714921951293945, | |
| "learning_rate": 2.994962216624685e-05, | |
| "loss": 0.6384, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 7.754816055297852, | |
| "learning_rate": 2.9937027707808566e-05, | |
| "loss": 0.5809, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 8.383195877075195, | |
| "learning_rate": 2.992443324937028e-05, | |
| "loss": 0.5908, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 6.874183654785156, | |
| "learning_rate": 2.9911838790931988e-05, | |
| "loss": 0.5422, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 7.232607841491699, | |
| "learning_rate": 2.9899244332493704e-05, | |
| "loss": 0.6751, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 8.095083236694336, | |
| "learning_rate": 2.9886649874055416e-05, | |
| "loss": 0.5845, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 10.181581497192383, | |
| "learning_rate": 2.987405541561713e-05, | |
| "loss": 0.4039, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 7.251787185668945, | |
| "learning_rate": 2.9861460957178844e-05, | |
| "loss": 0.5025, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 11.612983703613281, | |
| "learning_rate": 2.9848866498740553e-05, | |
| "loss": 0.658, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 13.866792678833008, | |
| "learning_rate": 2.9836272040302266e-05, | |
| "loss": 0.5777, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 9.340723037719727, | |
| "learning_rate": 2.982367758186398e-05, | |
| "loss": 0.6047, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 10.717991828918457, | |
| "learning_rate": 2.9811083123425694e-05, | |
| "loss": 0.5881, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 10.31982421875, | |
| "learning_rate": 2.9798488664987406e-05, | |
| "loss": 0.4501, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 9.57770824432373, | |
| "learning_rate": 2.978589420654912e-05, | |
| "loss": 0.4792, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 9.465398788452148, | |
| "learning_rate": 2.977329974811083e-05, | |
| "loss": 0.5718, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 16.856624603271484, | |
| "learning_rate": 2.9760705289672544e-05, | |
| "loss": 0.7295, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 11.629205703735352, | |
| "learning_rate": 2.974811083123426e-05, | |
| "loss": 0.5327, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 8.08765697479248, | |
| "learning_rate": 2.973551637279597e-05, | |
| "loss": 0.6436, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 11.187536239624023, | |
| "learning_rate": 2.972292191435768e-05, | |
| "loss": 0.5741, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 12.667988777160645, | |
| "learning_rate": 2.9710327455919397e-05, | |
| "loss": 0.5328, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 8.135946273803711, | |
| "learning_rate": 2.969773299748111e-05, | |
| "loss": 0.436, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 8.500922203063965, | |
| "learning_rate": 2.968513853904282e-05, | |
| "loss": 0.6446, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 8.38931941986084, | |
| "learning_rate": 2.9672544080604534e-05, | |
| "loss": 0.4106, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 7.952789783477783, | |
| "learning_rate": 2.9659949622166247e-05, | |
| "loss": 0.5269, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 9.8812255859375, | |
| "learning_rate": 2.964735516372796e-05, | |
| "loss": 0.5114, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 8.14013671875, | |
| "learning_rate": 2.9634760705289675e-05, | |
| "loss": 0.3894, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 10.428199768066406, | |
| "learning_rate": 2.9622166246851387e-05, | |
| "loss": 0.5967, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 9.175215721130371, | |
| "learning_rate": 2.9609571788413096e-05, | |
| "loss": 0.4978, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 8.572174072265625, | |
| "learning_rate": 2.9596977329974812e-05, | |
| "loss": 0.6002, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.942105293273926, | |
| "learning_rate": 2.9584382871536524e-05, | |
| "loss": 0.4905, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 8.92253303527832, | |
| "learning_rate": 2.9571788413098237e-05, | |
| "loss": 0.378, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 10.829577445983887, | |
| "learning_rate": 2.955919395465995e-05, | |
| "loss": 0.5745, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 7.60324764251709, | |
| "learning_rate": 2.9546599496221662e-05, | |
| "loss": 0.3207, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.186948776245117, | |
| "learning_rate": 2.9534005037783378e-05, | |
| "loss": 0.292, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 7.360220909118652, | |
| "learning_rate": 2.952141057934509e-05, | |
| "loss": 0.7034, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 13.63427734375, | |
| "learning_rate": 2.9508816120906802e-05, | |
| "loss": 0.4843, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 11.388775825500488, | |
| "learning_rate": 2.9496221662468515e-05, | |
| "loss": 0.4537, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 13.440176963806152, | |
| "learning_rate": 2.9483627204030227e-05, | |
| "loss": 0.624, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 13.364521026611328, | |
| "learning_rate": 2.947103274559194e-05, | |
| "loss": 0.3167, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 9.809293746948242, | |
| "learning_rate": 2.9458438287153656e-05, | |
| "loss": 0.5945, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 14.221176147460938, | |
| "learning_rate": 2.9445843828715368e-05, | |
| "loss": 0.3922, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 9.054633140563965, | |
| "learning_rate": 2.9433249370277077e-05, | |
| "loss": 0.4583, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 15.587240219116211, | |
| "learning_rate": 2.9420654911838793e-05, | |
| "loss": 0.3764, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 19.746070861816406, | |
| "learning_rate": 2.9408060453400505e-05, | |
| "loss": 0.3708, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 7.772696018218994, | |
| "learning_rate": 2.9395465994962218e-05, | |
| "loss": 0.3178, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 13.086981773376465, | |
| "learning_rate": 2.938287153652393e-05, | |
| "loss": 0.4907, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 8.691929817199707, | |
| "learning_rate": 2.9370277078085643e-05, | |
| "loss": 0.2839, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 9.099030494689941, | |
| "learning_rate": 2.9357682619647355e-05, | |
| "loss": 0.4787, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 17.202730178833008, | |
| "learning_rate": 2.934508816120907e-05, | |
| "loss": 0.5272, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 14.722758293151855, | |
| "learning_rate": 2.9332493702770783e-05, | |
| "loss": 0.3865, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 8.29536247253418, | |
| "learning_rate": 2.9319899244332492e-05, | |
| "loss": 0.6003, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 10.03609848022461, | |
| "learning_rate": 2.9307304785894208e-05, | |
| "loss": 0.5586, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 15.111539840698242, | |
| "learning_rate": 2.929471032745592e-05, | |
| "loss": 0.3587, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 10.219131469726562, | |
| "learning_rate": 2.9282115869017633e-05, | |
| "loss": 0.4541, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 9.36807918548584, | |
| "learning_rate": 2.926952141057935e-05, | |
| "loss": 0.448, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 12.47349739074707, | |
| "learning_rate": 2.9256926952141058e-05, | |
| "loss": 0.585, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 7.583310604095459, | |
| "learning_rate": 2.924433249370277e-05, | |
| "loss": 0.2274, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 13.765729904174805, | |
| "learning_rate": 2.9231738035264486e-05, | |
| "loss": 0.4415, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 22.002029418945312, | |
| "learning_rate": 2.92191435768262e-05, | |
| "loss": 0.4643, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 15.840831756591797, | |
| "learning_rate": 2.9206549118387908e-05, | |
| "loss": 0.9231, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 22.16878318786621, | |
| "learning_rate": 2.9193954659949623e-05, | |
| "loss": 0.6477, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 10.195225715637207, | |
| "learning_rate": 2.9181360201511336e-05, | |
| "loss": 0.4786, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 11.7352294921875, | |
| "learning_rate": 2.9168765743073048e-05, | |
| "loss": 0.4849, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 12.2020902633667, | |
| "learning_rate": 2.9156171284634764e-05, | |
| "loss": 0.3908, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 13.499194145202637, | |
| "learning_rate": 2.9143576826196473e-05, | |
| "loss": 0.6328, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 20.535781860351562, | |
| "learning_rate": 2.9130982367758185e-05, | |
| "loss": 0.6886, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 11.518393516540527, | |
| "learning_rate": 2.91183879093199e-05, | |
| "loss": 0.6269, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 10.17136001586914, | |
| "learning_rate": 2.9105793450881614e-05, | |
| "loss": 0.5077, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 6.502359390258789, | |
| "learning_rate": 2.9093198992443326e-05, | |
| "loss": 0.5269, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 16.268077850341797, | |
| "learning_rate": 2.908060453400504e-05, | |
| "loss": 0.6138, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 11.319491386413574, | |
| "learning_rate": 2.906801007556675e-05, | |
| "loss": 0.4758, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 7.1629319190979, | |
| "learning_rate": 2.9055415617128463e-05, | |
| "loss": 0.4717, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 17.689434051513672, | |
| "learning_rate": 2.904282115869018e-05, | |
| "loss": 0.6062, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 16.246288299560547, | |
| "learning_rate": 2.903022670025189e-05, | |
| "loss": 0.4756, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 12.266759872436523, | |
| "learning_rate": 2.90176322418136e-05, | |
| "loss": 0.4271, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 7.64720344543457, | |
| "learning_rate": 2.9005037783375317e-05, | |
| "loss": 0.5505, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.0152153968811035, | |
| "learning_rate": 2.899244332493703e-05, | |
| "loss": 0.4172, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 10.39532470703125, | |
| "learning_rate": 2.897984886649874e-05, | |
| "loss": 0.659, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 6.066320896148682, | |
| "learning_rate": 2.8967254408060454e-05, | |
| "loss": 0.5022, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 9.511946678161621, | |
| "learning_rate": 2.8954659949622166e-05, | |
| "loss": 0.3605, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 5.521829605102539, | |
| "learning_rate": 2.894206549118388e-05, | |
| "loss": 0.2804, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 6.357683181762695, | |
| "learning_rate": 2.8929471032745595e-05, | |
| "loss": 0.343, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 4.500522136688232, | |
| "learning_rate": 2.8916876574307307e-05, | |
| "loss": 0.2611, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 7.982411861419678, | |
| "learning_rate": 2.8904282115869016e-05, | |
| "loss": 0.291, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 5.47086763381958, | |
| "learning_rate": 2.8891687657430732e-05, | |
| "loss": 0.2171, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 6.41118860244751, | |
| "learning_rate": 2.8879093198992444e-05, | |
| "loss": 0.2698, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 7.404595375061035, | |
| "learning_rate": 2.8866498740554157e-05, | |
| "loss": 0.3499, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 9.283653259277344, | |
| "learning_rate": 2.885390428211587e-05, | |
| "loss": 0.3149, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 4.450091361999512, | |
| "learning_rate": 2.884130982367758e-05, | |
| "loss": 0.2221, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 13.71849536895752, | |
| "learning_rate": 2.8828715365239294e-05, | |
| "loss": 0.3377, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 5.049820423126221, | |
| "learning_rate": 2.881612090680101e-05, | |
| "loss": 0.1135, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 10.600882530212402, | |
| "learning_rate": 2.8803526448362722e-05, | |
| "loss": 0.313, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 9.384568214416504, | |
| "learning_rate": 2.879093198992443e-05, | |
| "loss": 0.4214, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 14.095718383789062, | |
| "learning_rate": 2.8778337531486147e-05, | |
| "loss": 0.3608, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 12.552332878112793, | |
| "learning_rate": 2.876574307304786e-05, | |
| "loss": 0.6839, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 12.589719772338867, | |
| "learning_rate": 2.8753148614609572e-05, | |
| "loss": 0.3692, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 10.545365333557129, | |
| "learning_rate": 2.8740554156171288e-05, | |
| "loss": 0.6336, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 4.529080390930176, | |
| "learning_rate": 2.8727959697732997e-05, | |
| "loss": 0.0726, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 10.031196594238281, | |
| "learning_rate": 2.871536523929471e-05, | |
| "loss": 0.5852, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 7.78066873550415, | |
| "learning_rate": 2.8702770780856425e-05, | |
| "loss": 0.3664, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 9.381865501403809, | |
| "learning_rate": 2.8690176322418137e-05, | |
| "loss": 0.6564, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 11.74254035949707, | |
| "learning_rate": 2.8677581863979846e-05, | |
| "loss": 0.4658, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 11.351922988891602, | |
| "learning_rate": 2.8664987405541562e-05, | |
| "loss": 0.5164, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 13.590174674987793, | |
| "learning_rate": 2.8652392947103275e-05, | |
| "loss": 0.5728, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 10.585956573486328, | |
| "learning_rate": 2.8639798488664987e-05, | |
| "loss": 0.2159, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 5.4417219161987305, | |
| "learning_rate": 2.8627204030226703e-05, | |
| "loss": 0.3464, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 8.342070579528809, | |
| "learning_rate": 2.8614609571788412e-05, | |
| "loss": 0.5985, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 6.407668590545654, | |
| "learning_rate": 2.8602015113350128e-05, | |
| "loss": 0.347, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 9.523656845092773, | |
| "learning_rate": 2.858942065491184e-05, | |
| "loss": 0.4596, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 7.752206325531006, | |
| "learning_rate": 2.8576826196473553e-05, | |
| "loss": 0.5771, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 10.980051040649414, | |
| "learning_rate": 2.8564231738035265e-05, | |
| "loss": 0.7089, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 7.721608638763428, | |
| "learning_rate": 2.8551637279596978e-05, | |
| "loss": 0.3957, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 5.920357704162598, | |
| "learning_rate": 2.853904282115869e-05, | |
| "loss": 0.3956, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 6.167881965637207, | |
| "learning_rate": 2.8526448362720406e-05, | |
| "loss": 0.2437, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 18.24216651916504, | |
| "learning_rate": 2.8513853904282118e-05, | |
| "loss": 0.6398, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 9.355070114135742, | |
| "learning_rate": 2.8501259445843827e-05, | |
| "loss": 0.7392, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 7.304947853088379, | |
| "learning_rate": 2.8488664987405543e-05, | |
| "loss": 0.5611, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.211945533752441, | |
| "learning_rate": 2.8476070528967256e-05, | |
| "loss": 0.3727, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 11.621819496154785, | |
| "learning_rate": 2.8463476070528968e-05, | |
| "loss": 0.5283, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 7.048833847045898, | |
| "learning_rate": 2.8450881612090684e-05, | |
| "loss": 0.6998, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 4.535017967224121, | |
| "learning_rate": 2.8438287153652393e-05, | |
| "loss": 0.431, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 4.307721138000488, | |
| "learning_rate": 2.8425692695214105e-05, | |
| "loss": 0.2508, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 7.520541191101074, | |
| "learning_rate": 2.841309823677582e-05, | |
| "loss": 0.2839, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 6.359663486480713, | |
| "learning_rate": 2.8400503778337533e-05, | |
| "loss": 0.2315, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5.880284786224365, | |
| "learning_rate": 2.8387909319899243e-05, | |
| "loss": 0.2354, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 14.193618774414062, | |
| "learning_rate": 2.837531486146096e-05, | |
| "loss": 0.7256, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 4.47697639465332, | |
| "learning_rate": 2.836272040302267e-05, | |
| "loss": 0.1521, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 8.092729568481445, | |
| "learning_rate": 2.8350125944584383e-05, | |
| "loss": 0.6047, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 7.471700668334961, | |
| "learning_rate": 2.83375314861461e-05, | |
| "loss": 0.3359, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 21.315391540527344, | |
| "learning_rate": 2.8324937027707808e-05, | |
| "loss": 0.6584, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 18.247236251831055, | |
| "learning_rate": 2.831234256926952e-05, | |
| "loss": 0.7327, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 13.097492218017578, | |
| "learning_rate": 2.8299748110831236e-05, | |
| "loss": 0.6226, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 5.948911190032959, | |
| "learning_rate": 2.828715365239295e-05, | |
| "loss": 0.3375, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 13.252030372619629, | |
| "learning_rate": 2.827455919395466e-05, | |
| "loss": 0.2128, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 10.255802154541016, | |
| "learning_rate": 2.8261964735516374e-05, | |
| "loss": 0.1734, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 7.478372097015381, | |
| "learning_rate": 2.8249370277078086e-05, | |
| "loss": 0.6492, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 10.903780937194824, | |
| "learning_rate": 2.82367758186398e-05, | |
| "loss": 0.9533, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 8.005816459655762, | |
| "learning_rate": 2.8224181360201514e-05, | |
| "loss": 0.3773, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 8.27034854888916, | |
| "learning_rate": 2.8211586901763223e-05, | |
| "loss": 0.6595, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 7.85991907119751, | |
| "learning_rate": 2.8198992443324936e-05, | |
| "loss": 0.3601, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 7.255641460418701, | |
| "learning_rate": 2.818639798488665e-05, | |
| "loss": 0.4024, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 14.83194351196289, | |
| "learning_rate": 2.8173803526448364e-05, | |
| "loss": 0.6188, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 5.861283779144287, | |
| "learning_rate": 2.8161209068010076e-05, | |
| "loss": 0.493, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 5.700588226318359, | |
| "learning_rate": 2.814861460957179e-05, | |
| "loss": 0.3955, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 13.155322074890137, | |
| "learning_rate": 2.81360201511335e-05, | |
| "loss": 0.4269, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 8.21208381652832, | |
| "learning_rate": 2.8123425692695214e-05, | |
| "loss": 0.4352, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 6.507381439208984, | |
| "learning_rate": 2.811083123425693e-05, | |
| "loss": 0.2584, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 9.123990058898926, | |
| "learning_rate": 2.8098236775818642e-05, | |
| "loss": 0.5207, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 10.706199645996094, | |
| "learning_rate": 2.808564231738035e-05, | |
| "loss": 0.4213, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 5.8628621101379395, | |
| "learning_rate": 2.8073047858942067e-05, | |
| "loss": 0.3451, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 7.158113956451416, | |
| "learning_rate": 2.806045340050378e-05, | |
| "loss": 0.3274, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 5.258652210235596, | |
| "learning_rate": 2.804785894206549e-05, | |
| "loss": 0.3959, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 3.823733329772949, | |
| "learning_rate": 2.8035264483627204e-05, | |
| "loss": 0.2523, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 8.84554386138916, | |
| "learning_rate": 2.8022670025188917e-05, | |
| "loss": 0.553, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.735053539276123, | |
| "learning_rate": 2.801007556675063e-05, | |
| "loss": 0.393, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 4.995819091796875, | |
| "learning_rate": 2.7997481108312345e-05, | |
| "loss": 0.4338, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 7.378593444824219, | |
| "learning_rate": 2.7984886649874057e-05, | |
| "loss": 0.613, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 6.643065452575684, | |
| "learning_rate": 2.7972292191435766e-05, | |
| "loss": 0.2731, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 9.81127643585205, | |
| "learning_rate": 2.7959697732997482e-05, | |
| "loss": 0.3793, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 10.082533836364746, | |
| "learning_rate": 2.7947103274559194e-05, | |
| "loss": 0.6018, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 5.660975933074951, | |
| "learning_rate": 2.7934508816120907e-05, | |
| "loss": 0.3733, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 12.785099983215332, | |
| "learning_rate": 2.7921914357682623e-05, | |
| "loss": 0.7492, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 8.646162986755371, | |
| "learning_rate": 2.7909319899244332e-05, | |
| "loss": 0.5155, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 7.465343952178955, | |
| "learning_rate": 2.7896725440806044e-05, | |
| "loss": 0.4195, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 6.121210098266602, | |
| "learning_rate": 2.788413098236776e-05, | |
| "loss": 0.3969, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 8.49669361114502, | |
| "learning_rate": 2.7871536523929472e-05, | |
| "loss": 0.4192, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 4.441570281982422, | |
| "learning_rate": 2.785894206549118e-05, | |
| "loss": 0.2789, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 6.6631245613098145, | |
| "learning_rate": 2.7846347607052897e-05, | |
| "loss": 0.4081, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 5.974854469299316, | |
| "learning_rate": 2.783375314861461e-05, | |
| "loss": 0.3799, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.523766040802002, | |
| "learning_rate": 2.7821158690176322e-05, | |
| "loss": 0.1571, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 6.361337184906006, | |
| "learning_rate": 2.7808564231738038e-05, | |
| "loss": 0.3327, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 8.463950157165527, | |
| "learning_rate": 2.7795969773299747e-05, | |
| "loss": 0.5044, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 13.052382469177246, | |
| "learning_rate": 2.778337531486146e-05, | |
| "loss": 0.4517, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 22.617353439331055, | |
| "learning_rate": 2.7770780856423175e-05, | |
| "loss": 1.0641, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 9.117269515991211, | |
| "learning_rate": 2.7758186397984888e-05, | |
| "loss": 0.3886, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 12.824740409851074, | |
| "learning_rate": 2.77455919395466e-05, | |
| "loss": 0.3126, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 7.019243240356445, | |
| "learning_rate": 2.7732997481108313e-05, | |
| "loss": 0.2975, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 7.184720039367676, | |
| "learning_rate": 2.7720403022670025e-05, | |
| "loss": 0.5372, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.207572937011719, | |
| "learning_rate": 2.770780856423174e-05, | |
| "loss": 0.4707, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 10.323878288269043, | |
| "learning_rate": 2.7695214105793453e-05, | |
| "loss": 0.4206, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.704578399658203, | |
| "learning_rate": 2.7682619647355162e-05, | |
| "loss": 0.4494, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 7.675583362579346, | |
| "learning_rate": 2.7670025188916878e-05, | |
| "loss": 0.2075, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.299776554107666, | |
| "learning_rate": 2.765743073047859e-05, | |
| "loss": 0.2851, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 8.457651138305664, | |
| "learning_rate": 2.7644836272040303e-05, | |
| "loss": 0.4121, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 5.211854934692383, | |
| "learning_rate": 2.763224181360202e-05, | |
| "loss": 0.1637, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 18.794628143310547, | |
| "learning_rate": 2.7619647355163728e-05, | |
| "loss": 0.4972, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 8.111665725708008, | |
| "learning_rate": 2.760705289672544e-05, | |
| "loss": 0.3703, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 14.346935272216797, | |
| "learning_rate": 2.7594458438287156e-05, | |
| "loss": 0.4615, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 4.972769260406494, | |
| "learning_rate": 2.758186397984887e-05, | |
| "loss": 0.1846, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 10.939284324645996, | |
| "learning_rate": 2.756926952141058e-05, | |
| "loss": 0.9461, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 18.650239944458008, | |
| "learning_rate": 2.7556675062972293e-05, | |
| "loss": 0.6751, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 21.654159545898438, | |
| "learning_rate": 2.7544080604534006e-05, | |
| "loss": 0.6841, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 20.820980072021484, | |
| "learning_rate": 2.7531486146095718e-05, | |
| "loss": 0.4773, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 22.232995986938477, | |
| "learning_rate": 2.7518891687657434e-05, | |
| "loss": 0.8045, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 24.284286499023438, | |
| "learning_rate": 2.7506297229219143e-05, | |
| "loss": 0.574, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 7.939319610595703, | |
| "learning_rate": 2.7493702770780855e-05, | |
| "loss": 0.2071, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 13.728672981262207, | |
| "learning_rate": 2.748110831234257e-05, | |
| "loss": 0.7237, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 8.86567497253418, | |
| "learning_rate": 2.7468513853904284e-05, | |
| "loss": 0.2757, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 8.1141357421875, | |
| "learning_rate": 2.7455919395465996e-05, | |
| "loss": 0.2384, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 11.418375968933105, | |
| "learning_rate": 2.744332493702771e-05, | |
| "loss": 0.7057, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 14.610424995422363, | |
| "learning_rate": 2.743073047858942e-05, | |
| "loss": 0.3707, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 8.905698776245117, | |
| "learning_rate": 2.7418136020151133e-05, | |
| "loss": 0.6057, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 20.418991088867188, | |
| "learning_rate": 2.740554156171285e-05, | |
| "loss": 0.8607, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 13.551456451416016, | |
| "learning_rate": 2.7392947103274562e-05, | |
| "loss": 0.532, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 13.535616874694824, | |
| "learning_rate": 2.738035264483627e-05, | |
| "loss": 0.6038, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 11.406067848205566, | |
| "learning_rate": 2.7367758186397987e-05, | |
| "loss": 0.6093, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 10.88662338256836, | |
| "learning_rate": 2.73551637279597e-05, | |
| "loss": 0.7387, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 17.733755111694336, | |
| "learning_rate": 2.734256926952141e-05, | |
| "loss": 0.7371, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 6.4716901779174805, | |
| "learning_rate": 2.7329974811083124e-05, | |
| "loss": 0.3157, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 3.8103296756744385, | |
| "learning_rate": 2.7317380352644836e-05, | |
| "loss": 0.2807, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.197389602661133, | |
| "learning_rate": 2.730478589420655e-05, | |
| "loss": 0.3373, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 6.760047435760498, | |
| "learning_rate": 2.7292191435768265e-05, | |
| "loss": 0.3521, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 5.889042854309082, | |
| "learning_rate": 2.7279596977329977e-05, | |
| "loss": 0.276, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 7.752958297729492, | |
| "learning_rate": 2.7267002518891686e-05, | |
| "loss": 0.5561, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 11.166560173034668, | |
| "learning_rate": 2.7254408060453402e-05, | |
| "loss": 0.3495, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 7.210260391235352, | |
| "learning_rate": 2.7241813602015114e-05, | |
| "loss": 0.3042, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.527043342590332, | |
| "learning_rate": 2.7229219143576827e-05, | |
| "loss": 0.2351, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 18.023122787475586, | |
| "learning_rate": 2.721662468513854e-05, | |
| "loss": 0.6469, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 10.936578750610352, | |
| "learning_rate": 2.720403022670025e-05, | |
| "loss": 0.5859, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 11.246559143066406, | |
| "learning_rate": 2.7191435768261964e-05, | |
| "loss": 0.4637, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 10.577720642089844, | |
| "learning_rate": 2.717884130982368e-05, | |
| "loss": 0.5009, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.079160690307617, | |
| "learning_rate": 2.7166246851385392e-05, | |
| "loss": 0.1622, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 7.402092933654785, | |
| "learning_rate": 2.71536523929471e-05, | |
| "loss": 0.2807, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 11.498400688171387, | |
| "learning_rate": 2.7141057934508817e-05, | |
| "loss": 0.7531, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 4.898861885070801, | |
| "learning_rate": 2.712846347607053e-05, | |
| "loss": 0.2006, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 10.955698013305664, | |
| "learning_rate": 2.7115869017632242e-05, | |
| "loss": 0.4176, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 9.794878005981445, | |
| "learning_rate": 2.7103274559193958e-05, | |
| "loss": 0.3908, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 15.19322395324707, | |
| "learning_rate": 2.7090680100755667e-05, | |
| "loss": 0.6566, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 12.016098022460938, | |
| "learning_rate": 2.707808564231738e-05, | |
| "loss": 0.4483, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.229908466339111, | |
| "learning_rate": 2.7065491183879095e-05, | |
| "loss": 0.4335, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 10.778432846069336, | |
| "learning_rate": 2.7052896725440807e-05, | |
| "loss": 0.2455, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 7.074155807495117, | |
| "learning_rate": 2.7040302267002517e-05, | |
| "loss": 0.211, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 6.830694675445557, | |
| "learning_rate": 2.7027707808564232e-05, | |
| "loss": 0.1891, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 8.83045768737793, | |
| "learning_rate": 2.7015113350125945e-05, | |
| "loss": 0.4168, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 8.931233406066895, | |
| "learning_rate": 2.7002518891687657e-05, | |
| "loss": 0.3596, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 11.842453002929688, | |
| "learning_rate": 2.6989924433249373e-05, | |
| "loss": 0.5041, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 6.133688926696777, | |
| "learning_rate": 2.6977329974811082e-05, | |
| "loss": 0.5432, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 8.509514808654785, | |
| "learning_rate": 2.6964735516372794e-05, | |
| "loss": 0.3315, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 8.522913932800293, | |
| "learning_rate": 2.695214105793451e-05, | |
| "loss": 0.3872, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 4.950671672821045, | |
| "learning_rate": 2.6939546599496223e-05, | |
| "loss": 0.2198, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 12.004826545715332, | |
| "learning_rate": 2.6926952141057935e-05, | |
| "loss": 0.4331, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 5.216823577880859, | |
| "learning_rate": 2.6914357682619648e-05, | |
| "loss": 0.12, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 4.667407512664795, | |
| "learning_rate": 2.690176322418136e-05, | |
| "loss": 0.3283, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 10.124615669250488, | |
| "learning_rate": 2.6889168765743072e-05, | |
| "loss": 0.188, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 16.799171447753906, | |
| "learning_rate": 2.6876574307304788e-05, | |
| "loss": 0.5253, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 11.693323135375977, | |
| "learning_rate": 2.6863979848866497e-05, | |
| "loss": 0.3584, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 9.2306489944458, | |
| "learning_rate": 2.685138539042821e-05, | |
| "loss": 0.2927, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 6.715793132781982, | |
| "learning_rate": 2.6838790931989926e-05, | |
| "loss": 0.3034, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 10.9847993850708, | |
| "learning_rate": 2.6826196473551638e-05, | |
| "loss": 0.3885, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 9.826873779296875, | |
| "learning_rate": 2.681360201511335e-05, | |
| "loss": 0.4347, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 16.903553009033203, | |
| "learning_rate": 2.6801007556675063e-05, | |
| "loss": 0.4324, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 10.097508430480957, | |
| "learning_rate": 2.6788413098236775e-05, | |
| "loss": 0.4163, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 7.872523307800293, | |
| "learning_rate": 2.677581863979849e-05, | |
| "loss": 0.5585, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 5.794745922088623, | |
| "learning_rate": 2.6763224181360204e-05, | |
| "loss": 0.3675, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 8.695060729980469, | |
| "learning_rate": 2.6750629722921916e-05, | |
| "loss": 0.2789, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 10.348055839538574, | |
| "learning_rate": 2.673803526448363e-05, | |
| "loss": 0.4214, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 11.409734725952148, | |
| "learning_rate": 2.672544080604534e-05, | |
| "loss": 0.7848, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 10.424241065979004, | |
| "learning_rate": 2.6712846347607053e-05, | |
| "loss": 0.5209, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 5.8873114585876465, | |
| "learning_rate": 2.670025188916877e-05, | |
| "loss": 0.2265, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 11.455771446228027, | |
| "learning_rate": 2.6687657430730478e-05, | |
| "loss": 0.7453, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 12.967717170715332, | |
| "learning_rate": 2.667506297229219e-05, | |
| "loss": 0.3653, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 8.101251602172852, | |
| "learning_rate": 2.6662468513853906e-05, | |
| "loss": 0.3649, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 10.239398956298828, | |
| "learning_rate": 2.664987405541562e-05, | |
| "loss": 0.4807, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 7.652868270874023, | |
| "learning_rate": 2.663727959697733e-05, | |
| "loss": 0.4622, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 7.320528030395508, | |
| "learning_rate": 2.6624685138539044e-05, | |
| "loss": 0.3544, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 5.603414535522461, | |
| "learning_rate": 2.6612090680100756e-05, | |
| "loss": 0.225, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 6.3062310218811035, | |
| "learning_rate": 2.659949622166247e-05, | |
| "loss": 0.29, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 10.358262062072754, | |
| "learning_rate": 2.6586901763224184e-05, | |
| "loss": 0.5102, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 5.093788146972656, | |
| "learning_rate": 2.6574307304785897e-05, | |
| "loss": 0.3955, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 12.661267280578613, | |
| "learning_rate": 2.6561712846347606e-05, | |
| "loss": 0.4287, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.933968544006348, | |
| "learning_rate": 2.654911838790932e-05, | |
| "loss": 0.3987, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 10.704058647155762, | |
| "learning_rate": 2.6536523929471034e-05, | |
| "loss": 0.3831, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 6.526610374450684, | |
| "learning_rate": 2.6523929471032746e-05, | |
| "loss": 0.3872, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 5.706236839294434, | |
| "learning_rate": 2.651133501259446e-05, | |
| "loss": 0.4442, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 7.010911464691162, | |
| "learning_rate": 2.649874055415617e-05, | |
| "loss": 0.403, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 5.128185749053955, | |
| "learning_rate": 2.6486146095717884e-05, | |
| "loss": 0.2681, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 9.200146675109863, | |
| "learning_rate": 2.64735516372796e-05, | |
| "loss": 0.4966, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 10.710809707641602, | |
| "learning_rate": 2.6460957178841312e-05, | |
| "loss": 0.3411, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 6.845553874969482, | |
| "learning_rate": 2.644836272040302e-05, | |
| "loss": 0.4416, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 6.162066459655762, | |
| "learning_rate": 2.6435768261964737e-05, | |
| "loss": 0.5407, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 11.713993072509766, | |
| "learning_rate": 2.642317380352645e-05, | |
| "loss": 0.9063, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 7.366119384765625, | |
| "learning_rate": 2.641057934508816e-05, | |
| "loss": 0.4344, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 7.175751209259033, | |
| "learning_rate": 2.6397984886649878e-05, | |
| "loss": 0.4297, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 14.455536842346191, | |
| "learning_rate": 2.6385390428211587e-05, | |
| "loss": 0.5623, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 7.098339080810547, | |
| "learning_rate": 2.63727959697733e-05, | |
| "loss": 0.2626, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 9.281996726989746, | |
| "learning_rate": 2.6360201511335015e-05, | |
| "loss": 0.4656, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 8.613694190979004, | |
| "learning_rate": 2.6347607052896727e-05, | |
| "loss": 0.5498, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 4.918963432312012, | |
| "learning_rate": 2.6335012594458436e-05, | |
| "loss": 0.4148, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 12.958881378173828, | |
| "learning_rate": 2.6322418136020152e-05, | |
| "loss": 0.8399, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 6.894360065460205, | |
| "learning_rate": 2.6309823677581865e-05, | |
| "loss": 0.3625, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 11.97496223449707, | |
| "learning_rate": 2.6297229219143577e-05, | |
| "loss": 0.6333, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 6.575906276702881, | |
| "learning_rate": 2.6284634760705293e-05, | |
| "loss": 0.3271, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 8.077478408813477, | |
| "learning_rate": 2.6272040302267002e-05, | |
| "loss": 0.4545, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 7.899360179901123, | |
| "learning_rate": 2.6259445843828714e-05, | |
| "loss": 0.4815, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 9.990239143371582, | |
| "learning_rate": 2.624685138539043e-05, | |
| "loss": 0.631, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 7.54863166809082, | |
| "learning_rate": 2.6234256926952142e-05, | |
| "loss": 0.5138, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 7.893070220947266, | |
| "learning_rate": 2.6221662468513855e-05, | |
| "loss": 0.2928, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 6.223887920379639, | |
| "learning_rate": 2.6209068010075567e-05, | |
| "loss": 0.3628, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 9.773299217224121, | |
| "learning_rate": 2.619647355163728e-05, | |
| "loss": 0.5439, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 4.369682788848877, | |
| "learning_rate": 2.6183879093198992e-05, | |
| "loss": 0.2298, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.92218542098999, | |
| "learning_rate": 2.6171284634760708e-05, | |
| "loss": 0.2903, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.855688571929932, | |
| "learning_rate": 2.6158690176322417e-05, | |
| "loss": 0.3738, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.964177131652832, | |
| "learning_rate": 2.614609571788413e-05, | |
| "loss": 0.373, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 8.245409965515137, | |
| "learning_rate": 2.6133501259445845e-05, | |
| "loss": 0.4829, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 8.923148155212402, | |
| "learning_rate": 2.6120906801007558e-05, | |
| "loss": 0.3148, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 9.844786643981934, | |
| "learning_rate": 2.610831234256927e-05, | |
| "loss": 0.4606, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 5.273078441619873, | |
| "learning_rate": 2.6095717884130983e-05, | |
| "loss": 0.2387, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 8.784069061279297, | |
| "learning_rate": 2.6083123425692695e-05, | |
| "loss": 0.3051, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 7.740212917327881, | |
| "learning_rate": 2.6070528967254407e-05, | |
| "loss": 0.2269, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 12.335386276245117, | |
| "learning_rate": 2.6057934508816123e-05, | |
| "loss": 0.4262, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.678969383239746, | |
| "learning_rate": 2.6045340050377836e-05, | |
| "loss": 0.1791, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 9.690211296081543, | |
| "learning_rate": 2.6032745591939545e-05, | |
| "loss": 0.2351, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 12.03903865814209, | |
| "learning_rate": 2.602015113350126e-05, | |
| "loss": 0.8051, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 11.595259666442871, | |
| "learning_rate": 2.6007556675062973e-05, | |
| "loss": 0.3681, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.8068230152130127, | |
| "learning_rate": 2.5994962216624685e-05, | |
| "loss": 0.3157, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.258709907531738, | |
| "learning_rate": 2.5982367758186398e-05, | |
| "loss": 0.1873, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 7.911970615386963, | |
| "learning_rate": 2.596977329974811e-05, | |
| "loss": 0.2601, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 13.647623062133789, | |
| "learning_rate": 2.5957178841309823e-05, | |
| "loss": 0.6636, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 14.344672203063965, | |
| "learning_rate": 2.594458438287154e-05, | |
| "loss": 0.5215, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 6.633630752563477, | |
| "learning_rate": 2.593198992443325e-05, | |
| "loss": 0.4419, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 11.353543281555176, | |
| "learning_rate": 2.5919395465994963e-05, | |
| "loss": 0.4572, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 8.904594421386719, | |
| "learning_rate": 2.5906801007556676e-05, | |
| "loss": 0.3291, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 10.205378532409668, | |
| "learning_rate": 2.5894206549118388e-05, | |
| "loss": 0.3371, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 8.3673095703125, | |
| "learning_rate": 2.5881612090680104e-05, | |
| "loss": 0.322, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 7.420979022979736, | |
| "learning_rate": 2.5869017632241813e-05, | |
| "loss": 0.3825, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 12.808219909667969, | |
| "learning_rate": 2.5856423173803526e-05, | |
| "loss": 0.5258, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 9.836726188659668, | |
| "learning_rate": 2.584382871536524e-05, | |
| "loss": 0.3549, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 6.072993755340576, | |
| "learning_rate": 2.5831234256926954e-05, | |
| "loss": 0.3907, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 20.400632858276367, | |
| "learning_rate": 2.5818639798488666e-05, | |
| "loss": 0.8538, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 9.215600967407227, | |
| "learning_rate": 2.580604534005038e-05, | |
| "loss": 0.3813, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 8.703184127807617, | |
| "learning_rate": 2.579345088161209e-05, | |
| "loss": 0.6626, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 10.990998268127441, | |
| "learning_rate": 2.5780856423173803e-05, | |
| "loss": 0.4793, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 11.248719215393066, | |
| "learning_rate": 2.576826196473552e-05, | |
| "loss": 0.3488, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 6.082599639892578, | |
| "learning_rate": 2.5755667506297232e-05, | |
| "loss": 0.4904, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 6.6478495597839355, | |
| "learning_rate": 2.574307304785894e-05, | |
| "loss": 0.535, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 8.529887199401855, | |
| "learning_rate": 2.5730478589420657e-05, | |
| "loss": 0.6424, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 7.008981704711914, | |
| "learning_rate": 2.571788413098237e-05, | |
| "loss": 0.2907, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 4.586373805999756, | |
| "learning_rate": 2.570528967254408e-05, | |
| "loss": 0.2298, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 6.62999963760376, | |
| "learning_rate": 2.5692695214105794e-05, | |
| "loss": 0.3508, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 9.748456954956055, | |
| "learning_rate": 2.5680100755667506e-05, | |
| "loss": 0.4157, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 6.978835582733154, | |
| "learning_rate": 2.566750629722922e-05, | |
| "loss": 0.3145, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 5.523429870605469, | |
| "learning_rate": 2.5654911838790935e-05, | |
| "loss": 0.3471, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 3.892453193664551, | |
| "learning_rate": 2.5642317380352647e-05, | |
| "loss": 0.2324, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 4.895789623260498, | |
| "learning_rate": 2.5629722921914356e-05, | |
| "loss": 0.3082, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 5.085398197174072, | |
| "learning_rate": 2.5617128463476072e-05, | |
| "loss": 0.4067, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 8.951998710632324, | |
| "learning_rate": 2.5604534005037784e-05, | |
| "loss": 0.4093, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 6.7257537841796875, | |
| "learning_rate": 2.5591939546599497e-05, | |
| "loss": 0.4623, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 9.630943298339844, | |
| "learning_rate": 2.5579345088161213e-05, | |
| "loss": 0.3686, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 7.153581142425537, | |
| "learning_rate": 2.556675062972292e-05, | |
| "loss": 0.468, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 13.286303520202637, | |
| "learning_rate": 2.5554156171284634e-05, | |
| "loss": 0.6772, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 7.783265113830566, | |
| "learning_rate": 2.554156171284635e-05, | |
| "loss": 0.2495, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 9.463069915771484, | |
| "learning_rate": 2.5528967254408062e-05, | |
| "loss": 0.393, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 7.801197052001953, | |
| "learning_rate": 2.551637279596977e-05, | |
| "loss": 0.2972, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 8.433084487915039, | |
| "learning_rate": 2.5503778337531487e-05, | |
| "loss": 0.4681, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.388092041015625, | |
| "learning_rate": 2.54911838790932e-05, | |
| "loss": 0.513, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.502450942993164, | |
| "learning_rate": 2.5478589420654912e-05, | |
| "loss": 0.1494, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 4.962391376495361, | |
| "learning_rate": 2.5465994962216628e-05, | |
| "loss": 0.1157, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 6.759737014770508, | |
| "learning_rate": 2.5453400503778337e-05, | |
| "loss": 0.1665, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 7.5388407707214355, | |
| "learning_rate": 2.544080604534005e-05, | |
| "loss": 0.3471, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 8.41981315612793, | |
| "learning_rate": 2.5428211586901765e-05, | |
| "loss": 0.4093, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 9.60474681854248, | |
| "learning_rate": 2.5415617128463477e-05, | |
| "loss": 0.576, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 6.068057537078857, | |
| "learning_rate": 2.540302267002519e-05, | |
| "loss": 0.203, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 4.424688339233398, | |
| "learning_rate": 2.5390428211586902e-05, | |
| "loss": 0.3454, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 11.978801727294922, | |
| "learning_rate": 2.5377833753148615e-05, | |
| "loss": 0.5684, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 6.773107051849365, | |
| "learning_rate": 2.5365239294710327e-05, | |
| "loss": 0.176, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 19.249605178833008, | |
| "learning_rate": 2.5352644836272043e-05, | |
| "loss": 0.586, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 15.166516304016113, | |
| "learning_rate": 2.5340050377833752e-05, | |
| "loss": 1.1046, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 7.654723167419434, | |
| "learning_rate": 2.5327455919395464e-05, | |
| "loss": 0.5403, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 18.302127838134766, | |
| "learning_rate": 2.531486146095718e-05, | |
| "loss": 0.9826, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 11.380560874938965, | |
| "learning_rate": 2.5302267002518893e-05, | |
| "loss": 0.8694, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 12.14782428741455, | |
| "learning_rate": 2.5289672544080605e-05, | |
| "loss": 0.6133, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 5.418036460876465, | |
| "learning_rate": 2.5277078085642318e-05, | |
| "loss": 0.3352, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 14.057409286499023, | |
| "learning_rate": 2.526448362720403e-05, | |
| "loss": 0.5738, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 6.224894046783447, | |
| "learning_rate": 2.5251889168765742e-05, | |
| "loss": 0.2167, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.565800189971924, | |
| "learning_rate": 2.5239294710327458e-05, | |
| "loss": 0.2301, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.572880744934082, | |
| "learning_rate": 2.522670025188917e-05, | |
| "loss": 0.2257, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 9.148365020751953, | |
| "learning_rate": 2.521410579345088e-05, | |
| "loss": 0.7268, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 5.175827980041504, | |
| "learning_rate": 2.5201511335012596e-05, | |
| "loss": 0.2781, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 10.978497505187988, | |
| "learning_rate": 2.5188916876574308e-05, | |
| "loss": 0.9962, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 6.896300792694092, | |
| "learning_rate": 2.517632241813602e-05, | |
| "loss": 0.7353, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 6.292906284332275, | |
| "learning_rate": 2.5163727959697733e-05, | |
| "loss": 0.3912, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 5.655113697052002, | |
| "learning_rate": 2.5151133501259445e-05, | |
| "loss": 0.2604, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 16.326297760009766, | |
| "learning_rate": 2.5138539042821158e-05, | |
| "loss": 1.0246, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 8.862824440002441, | |
| "learning_rate": 2.5125944584382874e-05, | |
| "loss": 0.287, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 12.7918119430542, | |
| "learning_rate": 2.5113350125944586e-05, | |
| "loss": 0.5495, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 8.119661331176758, | |
| "learning_rate": 2.5100755667506295e-05, | |
| "loss": 0.6277, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 7.078393459320068, | |
| "learning_rate": 2.508816120906801e-05, | |
| "loss": 0.4774, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 8.347442626953125, | |
| "learning_rate": 2.5075566750629723e-05, | |
| "loss": 0.6561, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 5.844377040863037, | |
| "learning_rate": 2.5062972292191436e-05, | |
| "loss": 0.3074, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 10.436945915222168, | |
| "learning_rate": 2.505037783375315e-05, | |
| "loss": 0.5096, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 7.574183464050293, | |
| "learning_rate": 2.503778337531486e-05, | |
| "loss": 0.2668, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 6.011397361755371, | |
| "learning_rate": 2.5025188916876573e-05, | |
| "loss": 0.4125, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.520245552062988, | |
| "learning_rate": 2.501259445843829e-05, | |
| "loss": 0.5079, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.188296794891357, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0709, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7905362776025237, | |
| "eval_f1": 0.7903191872567364, | |
| "eval_loss": 0.4741531312465668, | |
| "eval_precision": 0.7992890189359668, | |
| "eval_recall": 0.8039643127134329, | |
| "eval_runtime": 1566.2518, | |
| "eval_samples_per_second": 1.012, | |
| "eval_steps_per_second": 0.064, | |
| "step": 397 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 2382, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "total_flos": 833667378908160.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |