| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 16.93658536585366, | |
| "eval_steps": 500, | |
| "global_step": 136, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11707317073170732, | |
| "grad_norm": 1.7099019289016724, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.3897, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23414634146341465, | |
| "grad_norm": 1.738899827003479, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.4044, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.35121951219512193, | |
| "grad_norm": 1.7416936159133911, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.4046, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4682926829268293, | |
| "grad_norm": 1.671994686126709, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.3969, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5853658536585366, | |
| "grad_norm": 1.6600017547607422, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.3879, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7024390243902439, | |
| "grad_norm": 1.6655784845352173, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.4045, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.8195121951219512, | |
| "grad_norm": 1.4272384643554688, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3847, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.9365853658536586, | |
| "grad_norm": 1.362514615058899, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.3937, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1170731707317074, | |
| "grad_norm": 1.4144306182861328, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 0.7474, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.2341463414634146, | |
| "grad_norm": 0.6451680064201355, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.3747, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.3512195121951218, | |
| "grad_norm": 0.5170267224311829, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.3633, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.4682926829268292, | |
| "grad_norm": 1.1590852737426758, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.3621, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.5853658536585367, | |
| "grad_norm": 1.368847131729126, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 0.3669, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.7024390243902439, | |
| "grad_norm": 1.361088752746582, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3594, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.819512195121951, | |
| "grad_norm": 1.1211354732513428, | |
| "learning_rate": 9.998342337571566e-06, | |
| "loss": 0.3587, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.9365853658536585, | |
| "grad_norm": 0.9656141996383667, | |
| "learning_rate": 9.993370449424153e-06, | |
| "loss": 0.3662, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.1170731707317074, | |
| "grad_norm": 1.4052422046661377, | |
| "learning_rate": 9.985087632242634e-06, | |
| "loss": 0.7114, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.234146341463415, | |
| "grad_norm": 0.6288913488388062, | |
| "learning_rate": 9.973499378072947e-06, | |
| "loss": 0.3471, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.351219512195122, | |
| "grad_norm": 0.5390053987503052, | |
| "learning_rate": 9.958613370680507e-06, | |
| "loss": 0.3291, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.4682926829268292, | |
| "grad_norm": 0.46044138073921204, | |
| "learning_rate": 9.940439480455386e-06, | |
| "loss": 0.3249, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.5853658536585367, | |
| "grad_norm": 0.5259054899215698, | |
| "learning_rate": 9.918989757867584e-06, | |
| "loss": 0.3269, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.7024390243902436, | |
| "grad_norm": 0.5149078965187073, | |
| "learning_rate": 9.89427842547679e-06, | |
| "loss": 0.3143, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.819512195121951, | |
| "grad_norm": 0.5114086270332336, | |
| "learning_rate": 9.866321868501914e-06, | |
| "loss": 0.3324, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.9365853658536585, | |
| "grad_norm": 0.43192946910858154, | |
| "learning_rate": 9.835138623956603e-06, | |
| "loss": 0.3318, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.1170731707317074, | |
| "grad_norm": 0.7294260859489441, | |
| "learning_rate": 9.80074936835801e-06, | |
| "loss": 0.5944, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.234146341463415, | |
| "grad_norm": 0.344719260931015, | |
| "learning_rate": 9.763176904016914e-06, | |
| "loss": 0.3038, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.351219512195122, | |
| "grad_norm": 0.32829755544662476, | |
| "learning_rate": 9.722446143918307e-06, | |
| "loss": 0.3014, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.4682926829268292, | |
| "grad_norm": 0.3701910078525543, | |
| "learning_rate": 9.678584095202468e-06, | |
| "loss": 0.2873, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5853658536585367, | |
| "grad_norm": 0.3082481324672699, | |
| "learning_rate": 9.631619841257477e-06, | |
| "loss": 0.2773, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.7024390243902436, | |
| "grad_norm": 0.30135631561279297, | |
| "learning_rate": 9.581584522435025e-06, | |
| "loss": 0.2872, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.819512195121951, | |
| "grad_norm": 0.32238733768463135, | |
| "learning_rate": 9.528511315402358e-06, | |
| "loss": 0.2813, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.9365853658536585, | |
| "grad_norm": 0.29879140853881836, | |
| "learning_rate": 9.472435411143979e-06, | |
| "loss": 0.2819, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.117073170731707, | |
| "grad_norm": 0.5245116353034973, | |
| "learning_rate": 9.413393991627737e-06, | |
| "loss": 0.5332, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 4.234146341463415, | |
| "grad_norm": 0.22240440547466278, | |
| "learning_rate": 9.351426205150778e-06, | |
| "loss": 0.2488, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.351219512195122, | |
| "grad_norm": 0.2558436691761017, | |
| "learning_rate": 9.286573140381663e-06, | |
| "loss": 0.2538, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.46829268292683, | |
| "grad_norm": 0.2555961012840271, | |
| "learning_rate": 9.218877799115929e-06, | |
| "loss": 0.2484, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.585365853658536, | |
| "grad_norm": 0.2195068597793579, | |
| "learning_rate": 9.148385067763094e-06, | |
| "loss": 0.2556, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 4.702439024390244, | |
| "grad_norm": 0.22761856019496918, | |
| "learning_rate": 9.075141687584056e-06, | |
| "loss": 0.2399, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.819512195121951, | |
| "grad_norm": 0.2372802197933197, | |
| "learning_rate": 8.999196223698599e-06, | |
| "loss": 0.2531, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.9365853658536585, | |
| "grad_norm": 0.21613965928554535, | |
| "learning_rate": 8.920599032883553e-06, | |
| "loss": 0.2354, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.117073170731707, | |
| "grad_norm": 0.505303680896759, | |
| "learning_rate": 8.839402230183e-06, | |
| "loss": 0.4786, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 5.234146341463415, | |
| "grad_norm": 0.202455073595047, | |
| "learning_rate": 8.755659654352599e-06, | |
| "loss": 0.2088, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.351219512195122, | |
| "grad_norm": 0.2058672457933426, | |
| "learning_rate": 8.669426832160997e-06, | |
| "loss": 0.2202, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 5.46829268292683, | |
| "grad_norm": 0.20359627902507782, | |
| "learning_rate": 8.580760941571968e-06, | |
| "loss": 0.2222, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.585365853658536, | |
| "grad_norm": 0.21618157625198364, | |
| "learning_rate": 8.489720773831717e-06, | |
| "loss": 0.1921, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.702439024390244, | |
| "grad_norm": 0.21850977838039398, | |
| "learning_rate": 8.396366694486466e-06, | |
| "loss": 0.2011, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.819512195121951, | |
| "grad_norm": 0.198263481259346, | |
| "learning_rate": 8.30076060335616e-06, | |
| "loss": 0.1963, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.9365853658536585, | |
| "grad_norm": 0.20818056166172028, | |
| "learning_rate": 8.202965893490877e-06, | |
| "loss": 0.197, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.117073170731707, | |
| "grad_norm": 0.2016800194978714, | |
| "learning_rate": 8.103047409137114e-06, | |
| "loss": 0.3993, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 6.234146341463415, | |
| "grad_norm": 0.1838838905096054, | |
| "learning_rate": 8.001071402741843e-06, | |
| "loss": 0.1706, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.351219512195122, | |
| "grad_norm": 0.20501823723316193, | |
| "learning_rate": 7.897105491022819e-06, | |
| "loss": 0.1721, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.46829268292683, | |
| "grad_norm": 0.20208866894245148, | |
| "learning_rate": 7.791218610134324e-06, | |
| "loss": 0.1705, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.585365853658536, | |
| "grad_norm": 0.2018779218196869, | |
| "learning_rate": 7.683480969958005e-06, | |
| "loss": 0.1647, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 6.702439024390244, | |
| "grad_norm": 0.18786238133907318, | |
| "learning_rate": 7.5739640075491546e-06, | |
| "loss": 0.1668, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.819512195121951, | |
| "grad_norm": 0.18166756629943848, | |
| "learning_rate": 7.462740339769323e-06, | |
| "loss": 0.1576, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.9365853658536585, | |
| "grad_norm": 0.20427466928958893, | |
| "learning_rate": 7.349883715136601e-06, | |
| "loss": 0.1608, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 7.117073170731707, | |
| "grad_norm": 0.45569679141044617, | |
| "learning_rate": 7.235468964925571e-06, | |
| "loss": 0.2812, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 7.234146341463415, | |
| "grad_norm": 0.20992225408554077, | |
| "learning_rate": 7.119571953549305e-06, | |
| "loss": 0.1276, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 7.351219512195122, | |
| "grad_norm": 0.19525876641273499, | |
| "learning_rate": 7.002269528256334e-06, | |
| "loss": 0.1332, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.46829268292683, | |
| "grad_norm": 0.25107672810554504, | |
| "learning_rate": 6.883639468175926e-06, | |
| "loss": 0.1389, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.585365853658536, | |
| "grad_norm": 0.20671367645263672, | |
| "learning_rate": 6.763760432745475e-06, | |
| "loss": 0.1254, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 7.702439024390244, | |
| "grad_norm": 0.20864763855934143, | |
| "learning_rate": 6.6427119095541745e-06, | |
| "loss": 0.1284, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.819512195121951, | |
| "grad_norm": 0.20397327840328217, | |
| "learning_rate": 6.520574161637591e-06, | |
| "loss": 0.1284, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 7.9365853658536585, | |
| "grad_norm": 0.21467316150665283, | |
| "learning_rate": 6.397428174258048e-06, | |
| "loss": 0.125, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 8.117073170731707, | |
| "grad_norm": 0.4409783184528351, | |
| "learning_rate": 6.273355601206143e-06, | |
| "loss": 0.2228, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 8.234146341463415, | |
| "grad_norm": 0.16991958022117615, | |
| "learning_rate": 6.148438710658979e-06, | |
| "loss": 0.1076, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.351219512195122, | |
| "grad_norm": 0.2262849658727646, | |
| "learning_rate": 6.022760330631006e-06, | |
| "loss": 0.1007, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 8.46829268292683, | |
| "grad_norm": 0.31786417961120605, | |
| "learning_rate": 5.896403794053679e-06, | |
| "loss": 0.1037, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.585365853658537, | |
| "grad_norm": 0.18665149807929993, | |
| "learning_rate": 5.76945288352031e-06, | |
| "loss": 0.1005, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 8.702439024390245, | |
| "grad_norm": 0.18690058588981628, | |
| "learning_rate": 5.641991775732756e-06, | |
| "loss": 0.0941, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.819512195121952, | |
| "grad_norm": 0.24329355359077454, | |
| "learning_rate": 5.514104985686802e-06, | |
| "loss": 0.096, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 8.93658536585366, | |
| "grad_norm": 0.1969362050294876, | |
| "learning_rate": 5.385877310633233e-06, | |
| "loss": 0.0827, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 9.117073170731707, | |
| "grad_norm": 0.4472348690032959, | |
| "learning_rate": 5.257393773851733e-06, | |
| "loss": 0.195, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 9.234146341463415, | |
| "grad_norm": 0.16258855164051056, | |
| "learning_rate": 5.1287395682749444e-06, | |
| "loss": 0.0767, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 9.351219512195122, | |
| "grad_norm": 0.17854134738445282, | |
| "learning_rate": 5e-06, | |
| "loss": 0.073, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 9.46829268292683, | |
| "grad_norm": 0.2503032088279724, | |
| "learning_rate": 4.871260431725058e-06, | |
| "loss": 0.0783, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.585365853658537, | |
| "grad_norm": 0.22658348083496094, | |
| "learning_rate": 4.742606226148268e-06, | |
| "loss": 0.0742, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 9.702439024390245, | |
| "grad_norm": 0.1775130182504654, | |
| "learning_rate": 4.614122689366769e-06, | |
| "loss": 0.0707, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.819512195121952, | |
| "grad_norm": 0.2040383219718933, | |
| "learning_rate": 4.485895014313198e-06, | |
| "loss": 0.072, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 9.93658536585366, | |
| "grad_norm": 0.19907240569591522, | |
| "learning_rate": 4.358008224267245e-06, | |
| "loss": 0.0709, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.117073170731707, | |
| "grad_norm": 0.36260291934013367, | |
| "learning_rate": 4.230547116479691e-06, | |
| "loss": 0.1259, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 10.234146341463415, | |
| "grad_norm": 0.1743169128894806, | |
| "learning_rate": 4.103596205946323e-06, | |
| "loss": 0.0597, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 10.351219512195122, | |
| "grad_norm": 0.1686151772737503, | |
| "learning_rate": 3.977239669368998e-06, | |
| "loss": 0.0583, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 10.46829268292683, | |
| "grad_norm": 0.2038871943950653, | |
| "learning_rate": 3.851561289341023e-06, | |
| "loss": 0.0566, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.585365853658537, | |
| "grad_norm": 0.23157887160778046, | |
| "learning_rate": 3.726644398793857e-06, | |
| "loss": 0.0561, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.702439024390245, | |
| "grad_norm": 0.1856566071510315, | |
| "learning_rate": 3.6025718257419532e-06, | |
| "loss": 0.0555, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.819512195121952, | |
| "grad_norm": 0.15233968198299408, | |
| "learning_rate": 3.4794258383624115e-06, | |
| "loss": 0.0546, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 10.93658536585366, | |
| "grad_norm": 0.21109993755817413, | |
| "learning_rate": 3.3572880904458267e-06, | |
| "loss": 0.0546, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 11.117073170731707, | |
| "grad_norm": 0.5099601149559021, | |
| "learning_rate": 3.236239567254526e-06, | |
| "loss": 0.1046, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 11.234146341463415, | |
| "grad_norm": 0.21494165062904358, | |
| "learning_rate": 3.116360531824074e-06, | |
| "loss": 0.041, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 11.351219512195122, | |
| "grad_norm": 0.13625772297382355, | |
| "learning_rate": 2.997730471743667e-06, | |
| "loss": 0.0416, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 11.46829268292683, | |
| "grad_norm": 0.20177845656871796, | |
| "learning_rate": 2.880428046450697e-06, | |
| "loss": 0.0391, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 11.585365853658537, | |
| "grad_norm": 0.23105856776237488, | |
| "learning_rate": 2.7645310350744296e-06, | |
| "loss": 0.0467, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.702439024390245, | |
| "grad_norm": 0.24163837730884552, | |
| "learning_rate": 2.6501162848634023e-06, | |
| "loss": 0.0429, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.819512195121952, | |
| "grad_norm": 0.19448842108249664, | |
| "learning_rate": 2.537259660230679e-06, | |
| "loss": 0.0424, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 11.93658536585366, | |
| "grad_norm": 0.14827539026737213, | |
| "learning_rate": 2.426035992450848e-06, | |
| "loss": 0.0421, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 12.117073170731707, | |
| "grad_norm": 0.2161850482225418, | |
| "learning_rate": 2.316519030041998e-06, | |
| "loss": 0.0719, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 12.234146341463415, | |
| "grad_norm": 0.2117854356765747, | |
| "learning_rate": 2.2087813898656775e-06, | |
| "loss": 0.0359, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 12.351219512195122, | |
| "grad_norm": 0.16570013761520386, | |
| "learning_rate": 2.102894508977182e-06, | |
| "loss": 0.0331, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 12.46829268292683, | |
| "grad_norm": 0.11637657135725021, | |
| "learning_rate": 1.9989285972581595e-06, | |
| "loss": 0.0328, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.585365853658537, | |
| "grad_norm": 0.14085301756858826, | |
| "learning_rate": 1.896952590862886e-06, | |
| "loss": 0.033, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 12.702439024390245, | |
| "grad_norm": 0.23084422945976257, | |
| "learning_rate": 1.7970341065091246e-06, | |
| "loss": 0.0323, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.819512195121952, | |
| "grad_norm": 0.22168132662773132, | |
| "learning_rate": 1.699239396643841e-06, | |
| "loss": 0.0351, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 12.93658536585366, | |
| "grad_norm": 0.14602220058441162, | |
| "learning_rate": 1.6036333055135345e-06, | |
| "loss": 0.0302, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 13.117073170731707, | |
| "grad_norm": 0.2466672956943512, | |
| "learning_rate": 1.5102792261682813e-06, | |
| "loss": 0.0591, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 13.234146341463415, | |
| "grad_norm": 0.11887188255786896, | |
| "learning_rate": 1.4192390584280347e-06, | |
| "loss": 0.0301, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 13.351219512195122, | |
| "grad_norm": 0.14817075431346893, | |
| "learning_rate": 1.330573167839005e-06, | |
| "loss": 0.0295, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 13.46829268292683, | |
| "grad_norm": 0.12275740504264832, | |
| "learning_rate": 1.2443403456474017e-06, | |
| "loss": 0.0262, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 13.585365853658537, | |
| "grad_norm": 0.1150999441742897, | |
| "learning_rate": 1.1605977698170001e-06, | |
| "loss": 0.0247, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 13.702439024390245, | |
| "grad_norm": 0.10318116098642349, | |
| "learning_rate": 1.0794009671164484e-06, | |
| "loss": 0.0318, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.819512195121952, | |
| "grad_norm": 0.09050295501947403, | |
| "learning_rate": 1.0008037763014033e-06, | |
| "loss": 0.0249, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 13.93658536585366, | |
| "grad_norm": 0.09321045130491257, | |
| "learning_rate": 9.248583124159438e-07, | |
| "loss": 0.0255, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 14.117073170731707, | |
| "grad_norm": 0.21720270812511444, | |
| "learning_rate": 8.516149322369055e-07, | |
| "loss": 0.0466, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 14.234146341463415, | |
| "grad_norm": 0.08144959807395935, | |
| "learning_rate": 7.811222008840719e-07, | |
| "loss": 0.0228, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 14.351219512195122, | |
| "grad_norm": 0.08457339555025101, | |
| "learning_rate": 7.13426859618338e-07, | |
| "loss": 0.0237, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 14.46829268292683, | |
| "grad_norm": 0.09197661280632019, | |
| "learning_rate": 6.485737948492237e-07, | |
| "loss": 0.0246, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 14.585365853658537, | |
| "grad_norm": 0.09241025894880295, | |
| "learning_rate": 5.866060083722624e-07, | |
| "loss": 0.0231, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 14.702439024390245, | |
| "grad_norm": 0.09598127007484436, | |
| "learning_rate": 5.275645888560233e-07, | |
| "loss": 0.0237, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 14.819512195121952, | |
| "grad_norm": 0.11634089052677155, | |
| "learning_rate": 4.71488684597643e-07, | |
| "loss": 0.0231, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 14.93658536585366, | |
| "grad_norm": 0.08203331381082535, | |
| "learning_rate": 4.184154775649768e-07, | |
| "loss": 0.0219, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 15.117073170731707, | |
| "grad_norm": 0.18879924714565277, | |
| "learning_rate": 3.683801587425251e-07, | |
| "loss": 0.0445, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 15.234146341463415, | |
| "grad_norm": 0.07502248883247375, | |
| "learning_rate": 3.214159047975324e-07, | |
| "loss": 0.0205, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 15.351219512195122, | |
| "grad_norm": 0.0748981386423111, | |
| "learning_rate": 2.7755385608169374e-07, | |
| "loss": 0.0226, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 15.46829268292683, | |
| "grad_norm": 0.08773289620876312, | |
| "learning_rate": 2.368230959830875e-07, | |
| "loss": 0.0242, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 15.585365853658537, | |
| "grad_norm": 0.0750112384557724, | |
| "learning_rate": 1.992506316419912e-07, | |
| "loss": 0.0207, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 15.702439024390245, | |
| "grad_norm": 0.08113798499107361, | |
| "learning_rate": 1.6486137604339813e-07, | |
| "loss": 0.0212, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 15.819512195121952, | |
| "grad_norm": 0.08700235933065414, | |
| "learning_rate": 1.3367813149808728e-07, | |
| "loss": 0.0217, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 15.93658536585366, | |
| "grad_norm": 0.08342888951301575, | |
| "learning_rate": 1.0572157452321097e-07, | |
| "loss": 0.0223, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 16.117073170731707, | |
| "grad_norm": 0.18608692288398743, | |
| "learning_rate": 8.101024213241826e-08, | |
| "loss": 0.0419, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 16.234146341463415, | |
| "grad_norm": 0.08005540072917938, | |
| "learning_rate": 5.9560519544614725e-08, | |
| "loss": 0.0197, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 16.351219512195122, | |
| "grad_norm": 0.069722980260849, | |
| "learning_rate": 4.138662931949255e-08, | |
| "loss": 0.0206, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 16.46829268292683, | |
| "grad_norm": 0.08637074381113052, | |
| "learning_rate": 2.6500621927054716e-08, | |
| "loss": 0.0208, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 16.585365853658537, | |
| "grad_norm": 0.07552176713943481, | |
| "learning_rate": 1.4912367757366485e-08, | |
| "loss": 0.0211, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 16.702439024390245, | |
| "grad_norm": 0.08451703935861588, | |
| "learning_rate": 6.629550575847355e-09, | |
| "loss": 0.021, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 16.819512195121952, | |
| "grad_norm": 0.07880198210477829, | |
| "learning_rate": 1.657662428434792e-09, | |
| "loss": 0.0232, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 16.93658536585366, | |
| "grad_norm": 0.07944809645414352, | |
| "learning_rate": 0.0, | |
| "loss": 0.0211, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 16.93658536585366, | |
| "step": 136, | |
| "total_flos": 484618197794816.0, | |
| "train_loss": 0.16710591034087188, | |
| "train_runtime": 26752.9317, | |
| "train_samples_per_second": 0.521, | |
| "train_steps_per_second": 0.005 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 136, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 17, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 484618197794816.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |