| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.36133694670280037, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0018066847335140017, | |
| "grad_norm": 2.8337757951637657, | |
| "learning_rate": 0.0, | |
| "loss": 0.8251, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0036133694670280035, | |
| "grad_norm": 2.8815146894381005, | |
| "learning_rate": 1.7857142857142858e-07, | |
| "loss": 0.8284, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005420054200542005, | |
| "grad_norm": 2.888366333138549, | |
| "learning_rate": 3.5714285714285716e-07, | |
| "loss": 0.8432, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007226738934056007, | |
| "grad_norm": 2.8554157896663708, | |
| "learning_rate": 5.357142857142857e-07, | |
| "loss": 0.8398, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009033423667570008, | |
| "grad_norm": 2.7996302090063954, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.8303, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01084010840108401, | |
| "grad_norm": 2.758540000006854, | |
| "learning_rate": 8.928571428571429e-07, | |
| "loss": 0.8306, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.012646793134598013, | |
| "grad_norm": 2.7440712027687453, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 0.8129, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.014453477868112014, | |
| "grad_norm": 2.789670841276354, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.8217, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.016260162601626018, | |
| "grad_norm": 2.637271873271262, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.8192, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.018066847335140017, | |
| "grad_norm": 2.6274284830887247, | |
| "learning_rate": 1.6071428571428574e-06, | |
| "loss": 0.8146, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01987353206865402, | |
| "grad_norm": 2.2614743320388992, | |
| "learning_rate": 1.7857142857142859e-06, | |
| "loss": 0.8022, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.02168021680216802, | |
| "grad_norm": 2.1218891806221882, | |
| "learning_rate": 1.9642857142857144e-06, | |
| "loss": 0.8131, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.023486901535682024, | |
| "grad_norm": 2.1362019297809636, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.8047, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.025293586269196026, | |
| "grad_norm": 2.043181941163786, | |
| "learning_rate": 2.321428571428572e-06, | |
| "loss": 0.8006, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.02710027100271003, | |
| "grad_norm": 1.4654153096579285, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.7863, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.028906955736224028, | |
| "grad_norm": 1.424950643183691, | |
| "learning_rate": 2.6785714285714285e-06, | |
| "loss": 0.7794, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03071364046973803, | |
| "grad_norm": 1.371800886806853, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.7871, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.032520325203252036, | |
| "grad_norm": 1.2910993566284936, | |
| "learning_rate": 3.0357142857142856e-06, | |
| "loss": 0.7792, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03432700993676603, | |
| "grad_norm": 1.2099658145135412, | |
| "learning_rate": 3.2142857142857147e-06, | |
| "loss": 0.7728, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.036133694670280034, | |
| "grad_norm": 1.0479474035735314, | |
| "learning_rate": 3.3928571428571435e-06, | |
| "loss": 0.743, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037940379403794036, | |
| "grad_norm": 1.112276534792924, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.7403, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.03974706413730804, | |
| "grad_norm": 1.3035077079579298, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.7296, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04155374887082204, | |
| "grad_norm": 1.3695661288387202, | |
| "learning_rate": 3.928571428571429e-06, | |
| "loss": 0.7302, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.04336043360433604, | |
| "grad_norm": 1.2631605637070566, | |
| "learning_rate": 4.107142857142857e-06, | |
| "loss": 0.7359, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.045167118337850046, | |
| "grad_norm": 0.9577414796019008, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.713, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04697380307136405, | |
| "grad_norm": 0.7548293526053346, | |
| "learning_rate": 4.464285714285715e-06, | |
| "loss": 0.7275, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.04878048780487805, | |
| "grad_norm": 0.6708116039946823, | |
| "learning_rate": 4.642857142857144e-06, | |
| "loss": 0.7021, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.05058717253839205, | |
| "grad_norm": 0.7960055478682195, | |
| "learning_rate": 4.821428571428572e-06, | |
| "loss": 0.6998, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.052393857271906055, | |
| "grad_norm": 0.8790784871024999, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7086, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05420054200542006, | |
| "grad_norm": 0.9579979234005123, | |
| "learning_rate": 4.999955240022903e-06, | |
| "loss": 0.7065, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05600722673893405, | |
| "grad_norm": 1.003933265230055, | |
| "learning_rate": 4.999820961694372e-06, | |
| "loss": 0.6937, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.057813911472448055, | |
| "grad_norm": 0.6710805038763018, | |
| "learning_rate": 4.999597169822646e-06, | |
| "loss": 0.6812, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.05962059620596206, | |
| "grad_norm": 0.5001839964623817, | |
| "learning_rate": 4.999283872421259e-06, | |
| "loss": 0.6911, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.06142728093947606, | |
| "grad_norm": 0.5681434362593133, | |
| "learning_rate": 4.998881080708759e-06, | |
| "loss": 0.6766, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.06323396567299007, | |
| "grad_norm": 0.5752663504650961, | |
| "learning_rate": 4.998388809108304e-06, | |
| "loss": 0.6816, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06504065040650407, | |
| "grad_norm": 0.4990120714898939, | |
| "learning_rate": 4.997807075247147e-06, | |
| "loss": 0.6754, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.06684733514001806, | |
| "grad_norm": 0.4974519304946596, | |
| "learning_rate": 4.997135899956002e-06, | |
| "loss": 0.6782, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.06865401987353206, | |
| "grad_norm": 0.4349874019011085, | |
| "learning_rate": 4.996375307268303e-06, | |
| "loss": 0.6709, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07046070460704607, | |
| "grad_norm": 0.48312199739205014, | |
| "learning_rate": 4.995525324419338e-06, | |
| "loss": 0.6707, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07226738934056007, | |
| "grad_norm": 0.3760455090466998, | |
| "learning_rate": 4.994585981845278e-06, | |
| "loss": 0.6592, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 0.362711899175025, | |
| "learning_rate": 4.993557313182086e-06, | |
| "loss": 0.6666, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.07588075880758807, | |
| "grad_norm": 0.3527864788229306, | |
| "learning_rate": 4.992439355264308e-06, | |
| "loss": 0.656, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.07768744354110207, | |
| "grad_norm": 0.37039474310454246, | |
| "learning_rate": 4.9912321481237616e-06, | |
| "loss": 0.6567, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.07949412827461608, | |
| "grad_norm": 0.3381704985127657, | |
| "learning_rate": 4.989935734988098e-06, | |
| "loss": 0.6492, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08130081300813008, | |
| "grad_norm": 0.3368799245864057, | |
| "learning_rate": 4.988550162279255e-06, | |
| "loss": 0.6558, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08310749774164408, | |
| "grad_norm": 0.33217201284358455, | |
| "learning_rate": 4.9870754796117956e-06, | |
| "loss": 0.658, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.08491418247515808, | |
| "grad_norm": 0.3203999210132069, | |
| "learning_rate": 4.985511739791129e-06, | |
| "loss": 0.6596, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.08672086720867209, | |
| "grad_norm": 0.3108831280349791, | |
| "learning_rate": 4.983858998811622e-06, | |
| "loss": 0.6465, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.08852755194218609, | |
| "grad_norm": 0.28944557461245024, | |
| "learning_rate": 4.982117315854594e-06, | |
| "loss": 0.651, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09033423667570009, | |
| "grad_norm": 0.2724500694923891, | |
| "learning_rate": 4.980286753286196e-06, | |
| "loss": 0.6493, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0921409214092141, | |
| "grad_norm": 0.29233908216498683, | |
| "learning_rate": 4.978367376655177e-06, | |
| "loss": 0.6391, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0939476061427281, | |
| "grad_norm": 0.29050327837368234, | |
| "learning_rate": 4.976359254690543e-06, | |
| "loss": 0.6577, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0957542908762421, | |
| "grad_norm": 0.26277310012199956, | |
| "learning_rate": 4.974262459299088e-06, | |
| "loss": 0.641, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0975609756097561, | |
| "grad_norm": 0.2694317681957786, | |
| "learning_rate": 4.9720770655628216e-06, | |
| "loss": 0.6322, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0993676603432701, | |
| "grad_norm": 0.27643434259676186, | |
| "learning_rate": 4.969803151736285e-06, | |
| "loss": 0.6446, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1011743450767841, | |
| "grad_norm": 0.27938135134555137, | |
| "learning_rate": 4.967440799243739e-06, | |
| "loss": 0.6505, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.10298102981029811, | |
| "grad_norm": 0.2723254558437683, | |
| "learning_rate": 4.964990092676263e-06, | |
| "loss": 0.6358, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.10478771454381211, | |
| "grad_norm": 0.2476211836425744, | |
| "learning_rate": 4.962451119788709e-06, | |
| "loss": 0.6435, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.10659439927732611, | |
| "grad_norm": 0.2385367032602417, | |
| "learning_rate": 4.959823971496575e-06, | |
| "loss": 0.6289, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.10840108401084012, | |
| "grad_norm": 0.24895994998658083, | |
| "learning_rate": 4.957108741872736e-06, | |
| "loss": 0.6299, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1102077687443541, | |
| "grad_norm": 0.28129654057596637, | |
| "learning_rate": 4.954305528144085e-06, | |
| "loss": 0.6339, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1120144534778681, | |
| "grad_norm": 0.25373425749505557, | |
| "learning_rate": 4.9514144306880506e-06, | |
| "loss": 0.6393, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.11382113821138211, | |
| "grad_norm": 0.2583299024885213, | |
| "learning_rate": 4.948435553028994e-06, | |
| "loss": 0.6383, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.11562782294489611, | |
| "grad_norm": 0.24423529044993955, | |
| "learning_rate": 4.9453690018345144e-06, | |
| "loss": 0.6359, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.11743450767841011, | |
| "grad_norm": 0.24605558051142096, | |
| "learning_rate": 4.942214886911619e-06, | |
| "loss": 0.631, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11924119241192412, | |
| "grad_norm": 0.2548336350045181, | |
| "learning_rate": 4.938973321202799e-06, | |
| "loss": 0.6323, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.12104787714543812, | |
| "grad_norm": 0.32845268505362657, | |
| "learning_rate": 4.935644420781978e-06, | |
| "loss": 0.6229, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.12285456187895212, | |
| "grad_norm": 0.25203625601717533, | |
| "learning_rate": 4.932228304850363e-06, | |
| "loss": 0.6386, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.12466124661246612, | |
| "grad_norm": 0.2467239132093227, | |
| "learning_rate": 4.9287250957321685e-06, | |
| "loss": 0.6401, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.12646793134598014, | |
| "grad_norm": 0.28444822854929847, | |
| "learning_rate": 4.925134918870245e-06, | |
| "loss": 0.646, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12827461607949414, | |
| "grad_norm": 0.23105312295557145, | |
| "learning_rate": 4.921457902821578e-06, | |
| "loss": 0.6164, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.13008130081300814, | |
| "grad_norm": 0.23048405537510025, | |
| "learning_rate": 4.917694179252692e-06, | |
| "loss": 0.6342, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.13188798554652212, | |
| "grad_norm": 0.24150602333720247, | |
| "learning_rate": 4.9138438829349296e-06, | |
| "loss": 0.6332, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.13369467028003612, | |
| "grad_norm": 0.2462886765781574, | |
| "learning_rate": 4.909907151739634e-06, | |
| "loss": 0.6246, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.13550135501355012, | |
| "grad_norm": 0.2821908512766841, | |
| "learning_rate": 4.9058841266332005e-06, | |
| "loss": 0.6227, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13730803974706413, | |
| "grad_norm": 0.25172603846639535, | |
| "learning_rate": 4.901774951672041e-06, | |
| "loss": 0.6181, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.13911472448057813, | |
| "grad_norm": 0.23478283786140364, | |
| "learning_rate": 4.897579773997415e-06, | |
| "loss": 0.6123, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.14092140921409213, | |
| "grad_norm": 0.24524464756556244, | |
| "learning_rate": 4.893298743830168e-06, | |
| "loss": 0.625, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.14272809394760613, | |
| "grad_norm": 0.2557156779275681, | |
| "learning_rate": 4.8889320144653525e-06, | |
| "loss": 0.6347, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.14453477868112014, | |
| "grad_norm": 0.2501383477274707, | |
| "learning_rate": 4.884479742266731e-06, | |
| "loss": 0.6435, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14634146341463414, | |
| "grad_norm": 0.2777905831842133, | |
| "learning_rate": 4.879942086661185e-06, | |
| "loss": 0.6249, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.2311051834333071, | |
| "learning_rate": 4.875319210133004e-06, | |
| "loss": 0.6186, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.14995483288166214, | |
| "grad_norm": 0.24510662627866567, | |
| "learning_rate": 4.870611278218066e-06, | |
| "loss": 0.6301, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.15176151761517614, | |
| "grad_norm": 0.2457506431539522, | |
| "learning_rate": 4.865818459497911e-06, | |
| "loss": 0.6221, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.15356820234869015, | |
| "grad_norm": 0.23750506338642635, | |
| "learning_rate": 4.860940925593703e-06, | |
| "loss": 0.6238, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.15537488708220415, | |
| "grad_norm": 0.24583995522041222, | |
| "learning_rate": 4.855978851160088e-06, | |
| "loss": 0.6245, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.15718157181571815, | |
| "grad_norm": 0.24412234277510458, | |
| "learning_rate": 4.850932413878934e-06, | |
| "loss": 0.6266, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.15898825654923215, | |
| "grad_norm": 0.23692543465028104, | |
| "learning_rate": 4.845801794452978e-06, | |
| "loss": 0.6307, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.16079494128274616, | |
| "grad_norm": 0.24055583410087153, | |
| "learning_rate": 4.8405871765993435e-06, | |
| "loss": 0.6258, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.16260162601626016, | |
| "grad_norm": 0.25682681695753845, | |
| "learning_rate": 4.8352887470429726e-06, | |
| "loss": 0.6193, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16440831074977416, | |
| "grad_norm": 0.2451987238773762, | |
| "learning_rate": 4.829906695509934e-06, | |
| "loss": 0.6175, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.16621499548328816, | |
| "grad_norm": 0.33308910526145086, | |
| "learning_rate": 4.824441214720629e-06, | |
| "loss": 0.6207, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.16802168021680217, | |
| "grad_norm": 0.2599636527935745, | |
| "learning_rate": 4.8188925003828945e-06, | |
| "loss": 0.6275, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.16982836495031617, | |
| "grad_norm": 0.24262017713570547, | |
| "learning_rate": 4.813260751184992e-06, | |
| "loss": 0.6255, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.17163504968383017, | |
| "grad_norm": 0.24155086390799096, | |
| "learning_rate": 4.8075461687884935e-06, | |
| "loss": 0.6184, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.17344173441734417, | |
| "grad_norm": 0.2399226609650259, | |
| "learning_rate": 4.801748957821061e-06, | |
| "loss": 0.6164, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.17524841915085818, | |
| "grad_norm": 0.22903596103655702, | |
| "learning_rate": 4.795869325869117e-06, | |
| "loss": 0.6229, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.17705510388437218, | |
| "grad_norm": 0.2307755782529314, | |
| "learning_rate": 4.7899074834704165e-06, | |
| "loss": 0.6173, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.17886178861788618, | |
| "grad_norm": 0.24823755288051938, | |
| "learning_rate": 4.783863644106502e-06, | |
| "loss": 0.6292, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.18066847335140018, | |
| "grad_norm": 0.21658543875130712, | |
| "learning_rate": 4.777738024195065e-06, | |
| "loss": 0.6032, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18247515808491419, | |
| "grad_norm": 0.24020742361139732, | |
| "learning_rate": 4.771530843082187e-06, | |
| "loss": 0.6184, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.1842818428184282, | |
| "grad_norm": 0.23250154874953444, | |
| "learning_rate": 4.765242323034498e-06, | |
| "loss": 0.6373, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.1860885275519422, | |
| "grad_norm": 0.2515794157481243, | |
| "learning_rate": 4.7588726892312085e-06, | |
| "loss": 0.6162, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1878952122854562, | |
| "grad_norm": 0.2388452484683859, | |
| "learning_rate": 4.752422169756048e-06, | |
| "loss": 0.6147, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.1897018970189702, | |
| "grad_norm": 0.23451396083233195, | |
| "learning_rate": 4.7458909955891015e-06, | |
| "loss": 0.6066, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1915085817524842, | |
| "grad_norm": 0.23211345055411792, | |
| "learning_rate": 4.7392794005985324e-06, | |
| "loss": 0.6076, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.1933152664859982, | |
| "grad_norm": 0.2405071836183994, | |
| "learning_rate": 4.732587621532214e-06, | |
| "loss": 0.6139, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1951219512195122, | |
| "grad_norm": 0.238214343050957, | |
| "learning_rate": 4.7258158980092475e-06, | |
| "loss": 0.6125, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1969286359530262, | |
| "grad_norm": 0.23118377153943317, | |
| "learning_rate": 4.718964472511386e-06, | |
| "loss": 0.6224, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.1987353206865402, | |
| "grad_norm": 0.4227293348441868, | |
| "learning_rate": 4.712033590374346e-06, | |
| "loss": 0.604, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2005420054200542, | |
| "grad_norm": 0.22706445039114057, | |
| "learning_rate": 4.705023499779031e-06, | |
| "loss": 0.6231, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2023486901535682, | |
| "grad_norm": 0.21914294121010527, | |
| "learning_rate": 4.6979344517426345e-06, | |
| "loss": 0.6157, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.2041553748870822, | |
| "grad_norm": 0.2160022263814738, | |
| "learning_rate": 4.690766700109659e-06, | |
| "loss": 0.6193, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.20596205962059622, | |
| "grad_norm": 0.2304707253911858, | |
| "learning_rate": 4.683520501542825e-06, | |
| "loss": 0.6021, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.20776874435411022, | |
| "grad_norm": 0.24127382191408644, | |
| "learning_rate": 4.676196115513876e-06, | |
| "loss": 0.6186, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.20957542908762422, | |
| "grad_norm": 0.2406722048256909, | |
| "learning_rate": 4.668793804294294e-06, | |
| "loss": 0.6129, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.21138211382113822, | |
| "grad_norm": 0.24481652156673936, | |
| "learning_rate": 4.661313832945904e-06, | |
| "loss": 0.6228, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.21318879855465223, | |
| "grad_norm": 0.2283120728906989, | |
| "learning_rate": 4.653756469311381e-06, | |
| "loss": 0.6104, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.21499548328816623, | |
| "grad_norm": 0.2191260930120605, | |
| "learning_rate": 4.646121984004666e-06, | |
| "loss": 0.6015, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.21680216802168023, | |
| "grad_norm": 0.24306778049828862, | |
| "learning_rate": 4.638410650401267e-06, | |
| "loss": 0.6193, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2186088527551942, | |
| "grad_norm": 0.2269775698273598, | |
| "learning_rate": 4.630622744628478e-06, | |
| "loss": 0.6157, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.2204155374887082, | |
| "grad_norm": 0.22557834203825844, | |
| "learning_rate": 4.622758545555485e-06, | |
| "loss": 0.6205, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.2388627744876031, | |
| "learning_rate": 4.614818334783384e-06, | |
| "loss": 0.6117, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2240289069557362, | |
| "grad_norm": 0.33690846185965356, | |
| "learning_rate": 4.606802396635098e-06, | |
| "loss": 0.5973, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.22583559168925021, | |
| "grad_norm": 0.2559933256711994, | |
| "learning_rate": 4.598711018145193e-06, | |
| "loss": 0.6035, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.22764227642276422, | |
| "grad_norm": 0.23041365241908557, | |
| "learning_rate": 4.590544489049602e-06, | |
| "loss": 0.6167, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.22944896115627822, | |
| "grad_norm": 0.2386416733431471, | |
| "learning_rate": 4.582303101775249e-06, | |
| "loss": 0.6081, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.23125564588979222, | |
| "grad_norm": 0.2253681313258335, | |
| "learning_rate": 4.573987151429579e-06, | |
| "loss": 0.6075, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.23306233062330622, | |
| "grad_norm": 0.21952470403470545, | |
| "learning_rate": 4.565596935789987e-06, | |
| "loss": 0.6031, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.23486901535682023, | |
| "grad_norm": 0.24331315011786012, | |
| "learning_rate": 4.557132755293164e-06, | |
| "loss": 0.6219, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.23667570009033423, | |
| "grad_norm": 0.24835834109503124, | |
| "learning_rate": 4.54859491302433e-06, | |
| "loss": 0.6154, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.23848238482384823, | |
| "grad_norm": 0.22732916433505923, | |
| "learning_rate": 4.539983714706383e-06, | |
| "loss": 0.5945, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.24028906955736223, | |
| "grad_norm": 0.2181618946471109, | |
| "learning_rate": 4.531299468688956e-06, | |
| "loss": 0.61, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.24209575429087624, | |
| "grad_norm": 0.2259645647372355, | |
| "learning_rate": 4.522542485937369e-06, | |
| "loss": 0.6105, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.24390243902439024, | |
| "grad_norm": 0.2176244229453355, | |
| "learning_rate": 4.5137130800215025e-06, | |
| "loss": 0.6069, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.24570912375790424, | |
| "grad_norm": 0.2552195449531324, | |
| "learning_rate": 4.50481156710456e-06, | |
| "loss": 0.6113, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.24751580849141824, | |
| "grad_norm": 0.23421861438168898, | |
| "learning_rate": 4.495838265931754e-06, | |
| "loss": 0.5952, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.24932249322493225, | |
| "grad_norm": 0.21810921222578047, | |
| "learning_rate": 4.486793497818889e-06, | |
| "loss": 0.6121, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.25112917795844625, | |
| "grad_norm": 0.22773114029207175, | |
| "learning_rate": 4.477677586640854e-06, | |
| "loss": 0.6156, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.2529358626919603, | |
| "grad_norm": 0.23705077325515508, | |
| "learning_rate": 4.4684908588200305e-06, | |
| "loss": 0.6156, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.25474254742547425, | |
| "grad_norm": 0.24780177201776643, | |
| "learning_rate": 4.4592336433146e-06, | |
| "loss": 0.5953, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.2565492321589883, | |
| "grad_norm": 0.22862755103918672, | |
| "learning_rate": 4.449906271606766e-06, | |
| "loss": 0.616, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.25835591689250226, | |
| "grad_norm": 0.2404243241200089, | |
| "learning_rate": 4.440509077690883e-06, | |
| "loss": 0.607, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2601626016260163, | |
| "grad_norm": 0.22293770252772532, | |
| "learning_rate": 4.431042398061499e-06, | |
| "loss": 0.5965, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.26196928635953026, | |
| "grad_norm": 0.20972193775216136, | |
| "learning_rate": 4.421506571701305e-06, | |
| "loss": 0.6093, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.26377597109304424, | |
| "grad_norm": 0.23518600200676912, | |
| "learning_rate": 4.411901940068997e-06, | |
| "loss": 0.6112, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.26558265582655827, | |
| "grad_norm": 0.22746108225760642, | |
| "learning_rate": 4.402228847087046e-06, | |
| "loss": 0.6037, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.26738934056007224, | |
| "grad_norm": 0.24903963512865804, | |
| "learning_rate": 4.3924876391293915e-06, | |
| "loss": 0.6048, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.26919602529358627, | |
| "grad_norm": 0.23472947723463866, | |
| "learning_rate": 4.382678665009028e-06, | |
| "loss": 0.6016, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.27100271002710025, | |
| "grad_norm": 0.23431207097830484, | |
| "learning_rate": 4.372802275965521e-06, | |
| "loss": 0.61, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2728093947606143, | |
| "grad_norm": 0.22422800322111083, | |
| "learning_rate": 4.362858825652428e-06, | |
| "loss": 0.5949, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.27461607949412825, | |
| "grad_norm": 0.23561711836643753, | |
| "learning_rate": 4.352848670124637e-06, | |
| "loss": 0.5996, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.2764227642276423, | |
| "grad_norm": 0.21652376026883982, | |
| "learning_rate": 4.342772167825613e-06, | |
| "loss": 0.6009, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.27822944896115626, | |
| "grad_norm": 0.22742532543566119, | |
| "learning_rate": 4.332629679574566e-06, | |
| "loss": 0.6063, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.2800361336946703, | |
| "grad_norm": 0.3993451377830147, | |
| "learning_rate": 4.322421568553529e-06, | |
| "loss": 0.5961, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.28184281842818426, | |
| "grad_norm": 0.21468028516819, | |
| "learning_rate": 4.312148200294355e-06, | |
| "loss": 0.598, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.2836495031616983, | |
| "grad_norm": 0.2790705925726011, | |
| "learning_rate": 4.3018099426656255e-06, | |
| "loss": 0.6108, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.28545618789521227, | |
| "grad_norm": 0.22328718703580291, | |
| "learning_rate": 4.291407165859481e-06, | |
| "loss": 0.6073, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.2872628726287263, | |
| "grad_norm": 0.24079715236297133, | |
| "learning_rate": 4.280940242378363e-06, | |
| "loss": 0.5892, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.28906955736224027, | |
| "grad_norm": 0.3349816071529816, | |
| "learning_rate": 4.2704095470216745e-06, | |
| "loss": 0.5985, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2908762420957543, | |
| "grad_norm": 0.2351483485898414, | |
| "learning_rate": 4.259815456872363e-06, | |
| "loss": 0.6041, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.2926829268292683, | |
| "grad_norm": 0.23385513145667092, | |
| "learning_rate": 4.249158351283414e-06, | |
| "loss": 0.5903, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.2944896115627823, | |
| "grad_norm": 0.22962794810239154, | |
| "learning_rate": 4.2384386118642696e-06, | |
| "loss": 0.6006, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.227041646781334, | |
| "learning_rate": 4.227656622467162e-06, | |
| "loss": 0.609, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.2981029810298103, | |
| "grad_norm": 0.23339515173861125, | |
| "learning_rate": 4.216812769173371e-06, | |
| "loss": 0.6008, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2999096657633243, | |
| "grad_norm": 0.2279892093631049, | |
| "learning_rate": 4.205907440279395e-06, | |
| "loss": 0.6067, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3017163504968383, | |
| "grad_norm": 0.24349945347421037, | |
| "learning_rate": 4.194941026283053e-06, | |
| "loss": 0.6064, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3035230352303523, | |
| "grad_norm": 0.2191845217404836, | |
| "learning_rate": 4.183913919869495e-06, | |
| "loss": 0.6039, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3053297199638663, | |
| "grad_norm": 0.2180478057517194, | |
| "learning_rate": 4.172826515897146e-06, | |
| "loss": 0.597, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.3071364046973803, | |
| "grad_norm": 0.22148092498016186, | |
| "learning_rate": 4.161679211383565e-06, | |
| "loss": 0.6116, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3089430894308943, | |
| "grad_norm": 0.21976711391650572, | |
| "learning_rate": 4.150472405491226e-06, | |
| "loss": 0.5937, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.3107497741644083, | |
| "grad_norm": 0.2359272043525047, | |
| "learning_rate": 4.139206499513231e-06, | |
| "loss": 0.6002, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.31255645889792233, | |
| "grad_norm": 0.20743057651505534, | |
| "learning_rate": 4.127881896858934e-06, | |
| "loss": 0.6036, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3143631436314363, | |
| "grad_norm": 0.2185478761510027, | |
| "learning_rate": 4.116499003039499e-06, | |
| "loss": 0.5947, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.31616982836495033, | |
| "grad_norm": 0.24479063305435536, | |
| "learning_rate": 4.105058225653381e-06, | |
| "loss": 0.6, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3179765130984643, | |
| "grad_norm": 0.22541665254858562, | |
| "learning_rate": 4.093559974371725e-06, | |
| "loss": 0.6137, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.31978319783197834, | |
| "grad_norm": 0.23354779603092837, | |
| "learning_rate": 4.0820046609237026e-06, | |
| "loss": 0.5881, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.3215898825654923, | |
| "grad_norm": 0.22004453141007924, | |
| "learning_rate": 4.070392699081767e-06, | |
| "loss": 0.5948, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.32339656729900634, | |
| "grad_norm": 0.22301964541218125, | |
| "learning_rate": 4.058724504646834e-06, | |
| "loss": 0.6054, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3252032520325203, | |
| "grad_norm": 0.23125154097212539, | |
| "learning_rate": 4.047000495433397e-06, | |
| "loss": 0.6135, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32700993676603435, | |
| "grad_norm": 0.2194814968746533, | |
| "learning_rate": 4.035221091254563e-06, | |
| "loss": 0.5902, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.3288166214995483, | |
| "grad_norm": 0.22257729627885203, | |
| "learning_rate": 4.023386713907021e-06, | |
| "loss": 0.6005, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.33062330623306235, | |
| "grad_norm": 0.22264840616047507, | |
| "learning_rate": 4.011497787155938e-06, | |
| "loss": 0.6123, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.3324299909665763, | |
| "grad_norm": 0.22891861604269678, | |
| "learning_rate": 3.999554736719785e-06, | |
| "loss": 0.6093, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.33423667570009036, | |
| "grad_norm": 0.2368591129219394, | |
| "learning_rate": 3.987557990255093e-06, | |
| "loss": 0.6054, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.33604336043360433, | |
| "grad_norm": 0.2483359395611164, | |
| "learning_rate": 3.975507977341141e-06, | |
| "loss": 0.6011, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.33785004516711836, | |
| "grad_norm": 0.20257786753801169, | |
| "learning_rate": 3.963405129464569e-06, | |
| "loss": 0.5983, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.33965672990063234, | |
| "grad_norm": 0.2186759693689697, | |
| "learning_rate": 3.951249880003934e-06, | |
| "loss": 0.5973, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.34146341463414637, | |
| "grad_norm": 0.21876513682058207, | |
| "learning_rate": 3.939042664214185e-06, | |
| "loss": 0.6052, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.34327009936766034, | |
| "grad_norm": 0.2099916515113305, | |
| "learning_rate": 3.92678391921108e-06, | |
| "loss": 0.5936, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.34507678410117437, | |
| "grad_norm": 0.2343120602972887, | |
| "learning_rate": 3.914474083955537e-06, | |
| "loss": 0.592, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.34688346883468835, | |
| "grad_norm": 0.20255138476243217, | |
| "learning_rate": 3.902113599237911e-06, | |
| "loss": 0.5986, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3486901535682023, | |
| "grad_norm": 0.20455659875517468, | |
| "learning_rate": 3.889702907662212e-06, | |
| "loss": 0.6009, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.35049683830171635, | |
| "grad_norm": 0.22151992117493785, | |
| "learning_rate": 3.8772424536302565e-06, | |
| "loss": 0.6022, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3523035230352303, | |
| "grad_norm": 0.23833014531149102, | |
| "learning_rate": 3.864732683325755e-06, | |
| "loss": 0.5856, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.35411020776874436, | |
| "grad_norm": 0.2139415037008631, | |
| "learning_rate": 3.852174044698333e-06, | |
| "loss": 0.5989, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.35591689250225833, | |
| "grad_norm": 0.2194749540116677, | |
| "learning_rate": 3.839566987447492e-06, | |
| "loss": 0.598, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.35772357723577236, | |
| "grad_norm": 0.2381302592790127, | |
| "learning_rate": 3.826911963006508e-06, | |
| "loss": 0.5752, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.35953026196928634, | |
| "grad_norm": 0.23075119894415708, | |
| "learning_rate": 3.8142094245262617e-06, | |
| "loss": 0.5978, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.36133694670280037, | |
| "grad_norm": 0.23286158873769425, | |
| "learning_rate": 3.801459826859022e-06, | |
| "loss": 0.5877, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 553, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 231524290789376.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |