| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.0, | |
| "eval_steps": 500, | |
| "global_step": 231, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 6.3662924236000755, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 0.9017, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 6.470486978661982, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.8983, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 6.328464519368656, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9348, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 6.133977292798046, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.9185, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 4.744932065420312, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.8545, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 2.8651364128366312, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8065, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 2.4433934883263815, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.7942, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 3.782992999430219, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.7693, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 3.9199433485900963, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7663, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 3.929617185389812, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.7542, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 3.649295189975776, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.6991, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 3.1188753813712755, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6918, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 2.215664905854932, | |
| "learning_rate": 1.0833333333333334e-05, | |
| "loss": 0.635, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 1.866965377316569, | |
| "learning_rate": 1.1666666666666668e-05, | |
| "loss": 0.6434, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 2.235785768819185, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.6432, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 2.070505457914412, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.6213, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 1.6727932594181403, | |
| "learning_rate": 1.416666666666667e-05, | |
| "loss": 0.6167, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 1.1887337897149137, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.605, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 1.5143884211789567, | |
| "learning_rate": 1.5833333333333333e-05, | |
| "loss": 0.5892, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 1.3625293725217358, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.5958, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 1.0203660877207215, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 0.6375, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.1219569043143391, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.5816, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 0.9713461329458042, | |
| "learning_rate": 1.916666666666667e-05, | |
| "loss": 0.5923, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.7796387986969929, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5734, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 0.890524882817324, | |
| "learning_rate": 1.999884834944106e-05, | |
| "loss": 0.5786, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 0.7050988453366989, | |
| "learning_rate": 1.9995393663024054e-05, | |
| "loss": 0.5561, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.8513901497508272, | |
| "learning_rate": 1.9989636736467278e-05, | |
| "loss": 0.5631, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.6309950094973186, | |
| "learning_rate": 1.9981578895764272e-05, | |
| "loss": 0.5475, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 0.7377723459418936, | |
| "learning_rate": 1.9971221996878395e-05, | |
| "loss": 0.5293, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.7738449350922655, | |
| "learning_rate": 1.9958568425315316e-05, | |
| "loss": 0.5277, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 0.7266124322236066, | |
| "learning_rate": 1.9943621095573588e-05, | |
| "loss": 0.5138, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.6620476396452759, | |
| "learning_rate": 1.9926383450473344e-05, | |
| "loss": 0.5299, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.6903919450752821, | |
| "learning_rate": 1.9906859460363307e-05, | |
| "loss": 0.5355, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 0.7938178099559001, | |
| "learning_rate": 1.9885053622206305e-05, | |
| "loss": 0.4881, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.0606060606060606, | |
| "grad_norm": 0.7914815961889657, | |
| "learning_rate": 1.986097095854347e-05, | |
| "loss": 0.5088, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.6587783245311281, | |
| "learning_rate": 1.9834617016337424e-05, | |
| "loss": 0.4815, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.121212121212121, | |
| "grad_norm": 0.751200159055406, | |
| "learning_rate": 1.9805997865694616e-05, | |
| "loss": 0.4812, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 0.6727794828720169, | |
| "learning_rate": 1.9775120098467212e-05, | |
| "loss": 0.5023, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.1818181818181819, | |
| "grad_norm": 0.7531014479278245, | |
| "learning_rate": 1.9741990826734793e-05, | |
| "loss": 0.4687, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 0.7784381465344375, | |
| "learning_rate": 1.970661768116622e-05, | |
| "loss": 0.4926, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2424242424242424, | |
| "grad_norm": 0.6457063669854745, | |
| "learning_rate": 1.9669008809262064e-05, | |
| "loss": 0.4496, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.6780824336886752, | |
| "learning_rate": 1.9629172873477995e-05, | |
| "loss": 0.4361, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.303030303030303, | |
| "grad_norm": 0.7304649584581503, | |
| "learning_rate": 1.9587119049229558e-05, | |
| "loss": 0.4603, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.5557875051077491, | |
| "learning_rate": 1.954285702277879e-05, | |
| "loss": 0.4445, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.3636363636363638, | |
| "grad_norm": 0.6233262503300866, | |
| "learning_rate": 1.9496396989003195e-05, | |
| "loss": 0.4375, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.393939393939394, | |
| "grad_norm": 0.6013216394883922, | |
| "learning_rate": 1.944774964904754e-05, | |
| "loss": 0.4568, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.4242424242424243, | |
| "grad_norm": 0.6960796874829482, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 0.4492, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.6402361934677966, | |
| "learning_rate": 1.9343938371606714e-05, | |
| "loss": 0.4545, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.4848484848484849, | |
| "grad_norm": 0.7635621600679028, | |
| "learning_rate": 1.9288798344984673e-05, | |
| "loss": 0.4563, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.5151515151515151, | |
| "grad_norm": 0.7066501027260859, | |
| "learning_rate": 1.9231518828401458e-05, | |
| "loss": 0.4332, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5454545454545454, | |
| "grad_norm": 0.6684354794770975, | |
| "learning_rate": 1.917211301505453e-05, | |
| "loss": 0.4392, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.5757575757575757, | |
| "grad_norm": 0.7045261430123161, | |
| "learning_rate": 1.911059458789152e-05, | |
| "loss": 0.4373, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.606060606060606, | |
| "grad_norm": 0.7129812267235127, | |
| "learning_rate": 1.9046977716458627e-05, | |
| "loss": 0.456, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.779237200312619, | |
| "learning_rate": 1.8981277053636963e-05, | |
| "loss": 0.4569, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.6899018127267097, | |
| "learning_rate": 1.891350773226754e-05, | |
| "loss": 0.4297, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.696969696969697, | |
| "grad_norm": 0.6669601591538048, | |
| "learning_rate": 1.8843685361665724e-05, | |
| "loss": 0.4343, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.7272727272727273, | |
| "grad_norm": 0.7008920387240558, | |
| "learning_rate": 1.8771826024025944e-05, | |
| "loss": 0.4115, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.7575757575757576, | |
| "grad_norm": 0.60444752975965, | |
| "learning_rate": 1.8697946270717468e-05, | |
| "loss": 0.4435, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.7878787878787878, | |
| "grad_norm": 0.7133482116314053, | |
| "learning_rate": 1.8622063118472135e-05, | |
| "loss": 0.4784, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.6900103677023506, | |
| "learning_rate": 1.8544194045464888e-05, | |
| "loss": 0.4601, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8484848484848486, | |
| "grad_norm": 0.6089648948971949, | |
| "learning_rate": 1.8464356987288012e-05, | |
| "loss": 0.4295, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.878787878787879, | |
| "grad_norm": 0.6759016630249199, | |
| "learning_rate": 1.8382570332820045e-05, | |
| "loss": 0.4572, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 0.655706924741609, | |
| "learning_rate": 1.8298852919990254e-05, | |
| "loss": 0.4476, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.9393939393939394, | |
| "grad_norm": 0.5525937768604301, | |
| "learning_rate": 1.821322403143969e-05, | |
| "loss": 0.4559, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.9696969696969697, | |
| "grad_norm": 0.6080465832241057, | |
| "learning_rate": 1.812570339007983e-05, | |
| "loss": 0.4181, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5437215514964127, | |
| "learning_rate": 1.8036311154549783e-05, | |
| "loss": 0.4196, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.0303030303030303, | |
| "grad_norm": 0.7176035410218631, | |
| "learning_rate": 1.7945067914573147e-05, | |
| "loss": 0.3579, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.0606060606060606, | |
| "grad_norm": 0.6358460423195983, | |
| "learning_rate": 1.7851994686215592e-05, | |
| "loss": 0.3427, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.090909090909091, | |
| "grad_norm": 1.007260638281202, | |
| "learning_rate": 1.77571129070442e-05, | |
| "loss": 0.3602, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.121212121212121, | |
| "grad_norm": 0.591544667909819, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 0.3213, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.1515151515151514, | |
| "grad_norm": 0.9112853173299036, | |
| "learning_rate": 1.7562011524313187e-05, | |
| "loss": 0.3602, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.6346500398218357, | |
| "learning_rate": 1.7461836858476858e-05, | |
| "loss": 0.3169, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.212121212121212, | |
| "grad_norm": 0.6926405606826147, | |
| "learning_rate": 1.7359943506922775e-05, | |
| "loss": 0.3302, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.242424242424242, | |
| "grad_norm": 0.6452280160320382, | |
| "learning_rate": 1.725635493875799e-05, | |
| "loss": 0.3489, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 0.5840893157180582, | |
| "learning_rate": 1.7151095013548996e-05, | |
| "loss": 0.3427, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.303030303030303, | |
| "grad_norm": 0.6079541632641053, | |
| "learning_rate": 1.7044187975826126e-05, | |
| "loss": 0.3495, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.5971052759050507, | |
| "learning_rate": 1.693565844949933e-05, | |
| "loss": 0.3243, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.5946712776631644, | |
| "learning_rate": 1.6825531432186545e-05, | |
| "loss": 0.3261, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.393939393939394, | |
| "grad_norm": 0.5912304973085892, | |
| "learning_rate": 1.671383228945597e-05, | |
| "loss": 0.3313, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.4242424242424243, | |
| "grad_norm": 0.5922248085148802, | |
| "learning_rate": 1.6600586748983642e-05, | |
| "loss": 0.3634, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.4545454545454546, | |
| "grad_norm": 0.5796813201795245, | |
| "learning_rate": 1.648582089462756e-05, | |
| "loss": 0.3117, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.484848484848485, | |
| "grad_norm": 0.5872780500865236, | |
| "learning_rate": 1.6369561160419783e-05, | |
| "loss": 0.3529, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.515151515151515, | |
| "grad_norm": 0.5718922604221791, | |
| "learning_rate": 1.625183432447789e-05, | |
| "loss": 0.3141, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.5353007516484778, | |
| "learning_rate": 1.6132667502837164e-05, | |
| "loss": 0.3331, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.5757575757575757, | |
| "grad_norm": 0.5208464479347671, | |
| "learning_rate": 1.6012088143204953e-05, | |
| "loss": 0.3323, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.606060606060606, | |
| "grad_norm": 0.46108333913813826, | |
| "learning_rate": 1.589012401863864e-05, | |
| "loss": 0.3423, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.6363636363636362, | |
| "grad_norm": 0.6185954354592775, | |
| "learning_rate": 1.5766803221148676e-05, | |
| "loss": 0.3685, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.5232221010540287, | |
| "learning_rate": 1.5642154155228124e-05, | |
| "loss": 0.3321, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.6969696969696972, | |
| "grad_norm": 0.6040595615484364, | |
| "learning_rate": 1.5516205531310272e-05, | |
| "loss": 0.3161, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.48933287512749263, | |
| "learning_rate": 1.538898635915576e-05, | |
| "loss": 0.3205, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.757575757575758, | |
| "grad_norm": 0.5088971260831043, | |
| "learning_rate": 1.526052594117071e-05, | |
| "loss": 0.3471, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.787878787878788, | |
| "grad_norm": 0.5073220556688284, | |
| "learning_rate": 1.513085386565758e-05, | |
| "loss": 0.331, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.8181818181818183, | |
| "grad_norm": 0.530922230351584, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.3096, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.8484848484848486, | |
| "grad_norm": 0.48003003965710606, | |
| "learning_rate": 1.4867994483783485e-05, | |
| "loss": 0.3077, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.878787878787879, | |
| "grad_norm": 0.45380785530313666, | |
| "learning_rate": 1.4734867721853341e-05, | |
| "loss": 0.3251, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.5315535324111995, | |
| "learning_rate": 1.4600650377311523e-05, | |
| "loss": 0.3458, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.9393939393939394, | |
| "grad_norm": 0.48649781527198, | |
| "learning_rate": 1.4465373364454001e-05, | |
| "loss": 0.3476, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.9696969696969697, | |
| "grad_norm": 0.5607933022962, | |
| "learning_rate": 1.4329067841650274e-05, | |
| "loss": 0.3393, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.6083995531203883, | |
| "learning_rate": 1.4191765204166643e-05, | |
| "loss": 0.3153, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.0303030303030303, | |
| "grad_norm": 0.586034940697055, | |
| "learning_rate": 1.4053497076934948e-05, | |
| "loss": 0.2719, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.0606060606060606, | |
| "grad_norm": 0.5328898355116796, | |
| "learning_rate": 1.3914295307268396e-05, | |
| "loss": 0.2711, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.6512016566986433, | |
| "learning_rate": 1.3774191957526144e-05, | |
| "loss": 0.2362, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 3.121212121212121, | |
| "grad_norm": 0.6490664039207997, | |
| "learning_rate": 1.3633219297728415e-05, | |
| "loss": 0.2377, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 3.1515151515151514, | |
| "grad_norm": 0.5227982723425286, | |
| "learning_rate": 1.3491409798123687e-05, | |
| "loss": 0.2604, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 3.1818181818181817, | |
| "grad_norm": 0.6200307462962347, | |
| "learning_rate": 1.3348796121709862e-05, | |
| "loss": 0.2576, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 3.212121212121212, | |
| "grad_norm": 0.5936397387581552, | |
| "learning_rate": 1.3205411116710973e-05, | |
| "loss": 0.2482, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 3.242424242424242, | |
| "grad_norm": 0.5277502446070935, | |
| "learning_rate": 1.3061287809011243e-05, | |
| "loss": 0.2529, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.5654483028188876, | |
| "learning_rate": 1.291645939454825e-05, | |
| "loss": 0.2901, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 3.303030303030303, | |
| "grad_norm": 0.5351487800964878, | |
| "learning_rate": 1.277095923166689e-05, | |
| "loss": 0.2859, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.509220100035002, | |
| "learning_rate": 1.2624820833435939e-05, | |
| "loss": 0.236, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.3636363636363638, | |
| "grad_norm": 0.49327999676706624, | |
| "learning_rate": 1.2478077859929e-05, | |
| "loss": 0.2482, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 3.393939393939394, | |
| "grad_norm": 0.4580357431184566, | |
| "learning_rate": 1.2330764110471567e-05, | |
| "loss": 0.2449, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 3.4242424242424243, | |
| "grad_norm": 0.47910817823138163, | |
| "learning_rate": 1.2182913515856016e-05, | |
| "loss": 0.2496, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.4700731248939593, | |
| "learning_rate": 1.2034560130526341e-05, | |
| "loss": 0.2666, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 3.484848484848485, | |
| "grad_norm": 0.4429320491195175, | |
| "learning_rate": 1.1885738124734359e-05, | |
| "loss": 0.239, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 3.515151515151515, | |
| "grad_norm": 0.4570754510010221, | |
| "learning_rate": 1.1736481776669307e-05, | |
| "loss": 0.2825, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 3.5454545454545454, | |
| "grad_norm": 0.47154535576615286, | |
| "learning_rate": 1.1586825464562515e-05, | |
| "loss": 0.2571, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 3.5757575757575757, | |
| "grad_norm": 0.455331149252278, | |
| "learning_rate": 1.1436803658769082e-05, | |
| "loss": 0.2739, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.606060606060606, | |
| "grad_norm": 0.4459755731454009, | |
| "learning_rate": 1.1286450913828313e-05, | |
| "loss": 0.2413, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.44745867816587537, | |
| "learning_rate": 1.113580186050475e-05, | |
| "loss": 0.2538, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 0.4597570215046438, | |
| "learning_rate": 1.0984891197811686e-05, | |
| "loss": 0.2417, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.6969696969696972, | |
| "grad_norm": 0.42229239755840586, | |
| "learning_rate": 1.0833753685018935e-05, | |
| "loss": 0.2625, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.7272727272727275, | |
| "grad_norm": 0.43730649922839065, | |
| "learning_rate": 1.0682424133646712e-05, | |
| "loss": 0.2716, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.757575757575758, | |
| "grad_norm": 0.4437866415781284, | |
| "learning_rate": 1.0530937399447496e-05, | |
| "loss": 0.2285, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.787878787878788, | |
| "grad_norm": 0.43245220165630477, | |
| "learning_rate": 1.0379328374377715e-05, | |
| "loss": 0.2549, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.4404978612373396, | |
| "learning_rate": 1.0227631978561057e-05, | |
| "loss": 0.2656, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.8484848484848486, | |
| "grad_norm": 0.4572032796277578, | |
| "learning_rate": 1.0075883152245334e-05, | |
| "loss": 0.2287, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.878787878787879, | |
| "grad_norm": 0.42531491456929815, | |
| "learning_rate": 9.92411684775467e-06, | |
| "loss": 0.231, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.909090909090909, | |
| "grad_norm": 0.41287511795633053, | |
| "learning_rate": 9.772368021438943e-06, | |
| "loss": 0.263, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.9393939393939394, | |
| "grad_norm": 0.41599956248999415, | |
| "learning_rate": 9.620671625622287e-06, | |
| "loss": 0.2404, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.9696969696969697, | |
| "grad_norm": 0.42852471456183827, | |
| "learning_rate": 9.469062600552509e-06, | |
| "loss": 0.2349, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.43549516940285166, | |
| "learning_rate": 9.317575866353293e-06, | |
| "loss": 0.2536, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 4.03030303030303, | |
| "grad_norm": 0.4799073814984394, | |
| "learning_rate": 9.166246314981066e-06, | |
| "loss": 0.1892, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 4.0606060606060606, | |
| "grad_norm": 0.369420337018572, | |
| "learning_rate": 9.015108802188314e-06, | |
| "loss": 0.1967, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 4.090909090909091, | |
| "grad_norm": 0.4403498245451048, | |
| "learning_rate": 8.86419813949525e-06, | |
| "loss": 0.2053, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 4.121212121212121, | |
| "grad_norm": 0.5698228016958432, | |
| "learning_rate": 8.71354908617169e-06, | |
| "loss": 0.2107, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 4.151515151515151, | |
| "grad_norm": 0.42520797632350715, | |
| "learning_rate": 8.56319634123092e-06, | |
| "loss": 0.2142, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.4515877747433723, | |
| "learning_rate": 8.413174535437486e-06, | |
| "loss": 0.1994, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 4.212121212121212, | |
| "grad_norm": 0.43751298726966203, | |
| "learning_rate": 8.263518223330698e-06, | |
| "loss": 0.1712, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 4.242424242424242, | |
| "grad_norm": 0.44221396969712057, | |
| "learning_rate": 8.114261875265643e-06, | |
| "loss": 0.1919, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 4.2727272727272725, | |
| "grad_norm": 0.4253012458061858, | |
| "learning_rate": 7.965439869473664e-06, | |
| "loss": 0.2298, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 4.303030303030303, | |
| "grad_norm": 0.46392104648999866, | |
| "learning_rate": 7.817086484143987e-06, | |
| "loss": 0.2131, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 4.333333333333333, | |
| "grad_norm": 0.4191663951167429, | |
| "learning_rate": 7.669235889528436e-06, | |
| "loss": 0.1872, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.42771921281652686, | |
| "learning_rate": 7.521922140071003e-06, | |
| "loss": 0.23, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 4.393939393939394, | |
| "grad_norm": 0.40270934205581516, | |
| "learning_rate": 7.375179166564062e-06, | |
| "loss": 0.1873, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 4.424242424242424, | |
| "grad_norm": 0.38177718445068787, | |
| "learning_rate": 7.2290407683331154e-06, | |
| "loss": 0.1802, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 4.454545454545454, | |
| "grad_norm": 0.40715323350115823, | |
| "learning_rate": 7.0835406054517505e-06, | |
| "loss": 0.1633, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 4.484848484848484, | |
| "grad_norm": 0.37328460994016455, | |
| "learning_rate": 6.93871219098876e-06, | |
| "loss": 0.1948, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 4.515151515151516, | |
| "grad_norm": 0.36829523448149326, | |
| "learning_rate": 6.79458888328903e-06, | |
| "loss": 0.1988, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.37729635680302587, | |
| "learning_rate": 6.651203878290139e-06, | |
| "loss": 0.1839, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.575757575757576, | |
| "grad_norm": 0.3918622417901059, | |
| "learning_rate": 6.508590201876317e-06, | |
| "loss": 0.1959, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 4.606060606060606, | |
| "grad_norm": 0.3571046451175576, | |
| "learning_rate": 6.366780702271589e-06, | |
| "loss": 0.1771, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 4.636363636363637, | |
| "grad_norm": 0.4455431457067451, | |
| "learning_rate": 6.225808042473857e-06, | |
| "loss": 0.2182, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 0.3802179637239605, | |
| "learning_rate": 6.085704692731609e-06, | |
| "loss": 0.2084, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 4.696969696969697, | |
| "grad_norm": 0.3777379794122002, | |
| "learning_rate": 5.946502923065054e-06, | |
| "loss": 0.199, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.40376985012895866, | |
| "learning_rate": 5.8082347958333625e-06, | |
| "loss": 0.1736, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 4.757575757575758, | |
| "grad_norm": 0.3840163353104601, | |
| "learning_rate": 5.670932158349732e-06, | |
| "loss": 0.1885, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 4.787878787878788, | |
| "grad_norm": 0.40581108175379177, | |
| "learning_rate": 5.534626635546e-06, | |
| "loss": 0.2012, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.818181818181818, | |
| "grad_norm": 0.3885767820946026, | |
| "learning_rate": 5.399349622688479e-06, | |
| "loss": 0.2178, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.848484848484849, | |
| "grad_norm": 0.3785412326872522, | |
| "learning_rate": 5.2651322781466606e-06, | |
| "loss": 0.1998, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.878787878787879, | |
| "grad_norm": 0.38381401081938543, | |
| "learning_rate": 5.132005516216512e-06, | |
| "loss": 0.1812, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.36966836368030004, | |
| "learning_rate": 5.000000000000003e-06, | |
| "loss": 0.1942, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 4.9393939393939394, | |
| "grad_norm": 0.37169037028742086, | |
| "learning_rate": 4.869146134342426e-06, | |
| "loss": 0.1721, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 4.96969696969697, | |
| "grad_norm": 0.37264459388390875, | |
| "learning_rate": 4.739474058829288e-06, | |
| "loss": 0.2006, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.35711989972419017, | |
| "learning_rate": 4.611013640844245e-06, | |
| "loss": 0.2207, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 5.03030303030303, | |
| "grad_norm": 0.4249789146837776, | |
| "learning_rate": 4.483794468689728e-06, | |
| "loss": 0.1495, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 5.0606060606060606, | |
| "grad_norm": 0.3780102006420505, | |
| "learning_rate": 4.357845844771881e-06, | |
| "loss": 0.1617, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.3080141432769558, | |
| "learning_rate": 4.2331967788513295e-06, | |
| "loss": 0.1475, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 5.121212121212121, | |
| "grad_norm": 0.3693677715758717, | |
| "learning_rate": 4.109875981361363e-06, | |
| "loss": 0.1765, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 5.151515151515151, | |
| "grad_norm": 0.40258520208084353, | |
| "learning_rate": 3.987911856795047e-06, | |
| "loss": 0.1823, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 5.181818181818182, | |
| "grad_norm": 0.42647755216333316, | |
| "learning_rate": 3.867332497162836e-06, | |
| "loss": 0.1839, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 5.212121212121212, | |
| "grad_norm": 0.34439863538562254, | |
| "learning_rate": 3.748165675522113e-06, | |
| "loss": 0.1469, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 5.242424242424242, | |
| "grad_norm": 0.35087337270373725, | |
| "learning_rate": 3.630438839580217e-06, | |
| "loss": 0.1623, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.3259681077779499, | |
| "learning_rate": 3.5141791053724405e-06, | |
| "loss": 0.1757, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 5.303030303030303, | |
| "grad_norm": 0.38034914279283705, | |
| "learning_rate": 3.399413251016359e-06, | |
| "loss": 0.1766, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.333333333333333, | |
| "grad_norm": 0.3329709394838678, | |
| "learning_rate": 3.2861677105440335e-06, | |
| "loss": 0.1793, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 5.363636363636363, | |
| "grad_norm": 0.3518593635224805, | |
| "learning_rate": 3.174468567813461e-06, | |
| "loss": 0.1425, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 5.393939393939394, | |
| "grad_norm": 0.3388579551057604, | |
| "learning_rate": 3.0643415505006733e-06, | |
| "loss": 0.1904, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 5.424242424242424, | |
| "grad_norm": 0.31292443123856695, | |
| "learning_rate": 2.9558120241738786e-06, | |
| "loss": 0.1361, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.3298508190383936, | |
| "learning_rate": 2.8489049864510053e-06, | |
| "loss": 0.1612, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 5.484848484848484, | |
| "grad_norm": 0.30283528380823616, | |
| "learning_rate": 2.7436450612420098e-06, | |
| "loss": 0.168, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 5.515151515151516, | |
| "grad_norm": 0.2908061778716274, | |
| "learning_rate": 2.640056493077231e-06, | |
| "loss": 0.1115, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 5.545454545454545, | |
| "grad_norm": 0.32240181879624513, | |
| "learning_rate": 2.5381631415231455e-06, | |
| "loss": 0.1643, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 5.575757575757576, | |
| "grad_norm": 0.3236813038763638, | |
| "learning_rate": 2.4379884756868167e-06, | |
| "loss": 0.1812, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 5.606060606060606, | |
| "grad_norm": 0.3078959980344758, | |
| "learning_rate": 2.339555568810221e-06, | |
| "loss": 0.1491, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.282408685066355, | |
| "learning_rate": 2.2428870929558012e-06, | |
| "loss": 0.1814, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 5.666666666666667, | |
| "grad_norm": 0.31439301083057, | |
| "learning_rate": 2.1480053137844115e-06, | |
| "loss": 0.1455, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 5.696969696969697, | |
| "grad_norm": 0.2952494817710859, | |
| "learning_rate": 2.054932085426856e-06, | |
| "loss": 0.1718, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 5.7272727272727275, | |
| "grad_norm": 0.2993759221102418, | |
| "learning_rate": 1.963688845450218e-06, | |
| "loss": 0.1865, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 5.757575757575758, | |
| "grad_norm": 0.3085670439589066, | |
| "learning_rate": 1.8742966099201699e-06, | |
| "loss": 0.1654, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.787878787878788, | |
| "grad_norm": 0.31319957164367807, | |
| "learning_rate": 1.7867759685603115e-06, | |
| "loss": 0.1642, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.28158153804276187, | |
| "learning_rate": 1.7011470800097496e-06, | |
| "loss": 0.1495, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 5.848484848484849, | |
| "grad_norm": 0.31408467678087143, | |
| "learning_rate": 1.6174296671799571e-06, | |
| "loss": 0.1527, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 5.878787878787879, | |
| "grad_norm": 0.28475123222172366, | |
| "learning_rate": 1.5356430127119915e-06, | |
| "loss": 0.1477, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 5.909090909090909, | |
| "grad_norm": 0.2896881839307956, | |
| "learning_rate": 1.4558059545351144e-06, | |
| "loss": 0.1456, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 5.9393939393939394, | |
| "grad_norm": 0.29710799670787763, | |
| "learning_rate": 1.3779368815278648e-06, | |
| "loss": 0.186, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 5.96969696969697, | |
| "grad_norm": 0.27862714683095396, | |
| "learning_rate": 1.302053729282533e-06, | |
| "loss": 0.1437, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.3281693739817494, | |
| "learning_rate": 1.2281739759740575e-06, | |
| "loss": 0.1695, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 6.03030303030303, | |
| "grad_norm": 0.3165312644094072, | |
| "learning_rate": 1.156314638334277e-06, | |
| "loss": 0.1608, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 6.0606060606060606, | |
| "grad_norm": 0.2711138818846315, | |
| "learning_rate": 1.086492267732462e-06, | |
| "loss": 0.1757, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.090909090909091, | |
| "grad_norm": 0.2970179925615467, | |
| "learning_rate": 1.01872294636304e-06, | |
| "loss": 0.1293, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 6.121212121212121, | |
| "grad_norm": 0.24749289702225669, | |
| "learning_rate": 9.530222835413739e-07, | |
| "loss": 0.1264, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 6.151515151515151, | |
| "grad_norm": 0.24166485132175802, | |
| "learning_rate": 8.894054121084839e-07, | |
| "loss": 0.1559, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.2469176222804121, | |
| "learning_rate": 8.278869849454718e-07, | |
| "loss": 0.169, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 6.212121212121212, | |
| "grad_norm": 0.26389395222646517, | |
| "learning_rate": 7.684811715985429e-07, | |
| "loss": 0.1504, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 6.242424242424242, | |
| "grad_norm": 0.2740989338034577, | |
| "learning_rate": 7.1120165501533e-07, | |
| "loss": 0.1603, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 6.2727272727272725, | |
| "grad_norm": 0.294501191364033, | |
| "learning_rate": 6.560616283932897e-07, | |
| "loss": 0.1582, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 6.303030303030303, | |
| "grad_norm": 0.29335104020185554, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.1864, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 6.333333333333333, | |
| "grad_norm": 0.228840329850511, | |
| "learning_rate": 5.522503509524591e-07, | |
| "loss": 0.1079, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.28108871805858576, | |
| "learning_rate": 5.036030109968082e-07, | |
| "loss": 0.141, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 6.393939393939394, | |
| "grad_norm": 0.2573272348540072, | |
| "learning_rate": 4.5714297722121105e-07, | |
| "loss": 0.1627, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 6.424242424242424, | |
| "grad_norm": 0.2535362623881419, | |
| "learning_rate": 4.128809507704445e-07, | |
| "loss": 0.1536, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 6.454545454545454, | |
| "grad_norm": 0.23989911453624135, | |
| "learning_rate": 3.708271265220087e-07, | |
| "loss": 0.1259, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 6.484848484848484, | |
| "grad_norm": 0.24688340998236424, | |
| "learning_rate": 3.309911907379393e-07, | |
| "loss": 0.1587, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 6.515151515151516, | |
| "grad_norm": 0.2444755883515435, | |
| "learning_rate": 2.9338231883378365e-07, | |
| "loss": 0.1209, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.24171561212687892, | |
| "learning_rate": 2.5800917326521013e-07, | |
| "loss": 0.1364, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 6.575757575757576, | |
| "grad_norm": 0.22685601013938037, | |
| "learning_rate": 2.248799015327907e-07, | |
| "loss": 0.1175, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 6.606060606060606, | |
| "grad_norm": 0.2251854002494291, | |
| "learning_rate": 1.9400213430538773e-07, | |
| "loss": 0.1392, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 6.636363636363637, | |
| "grad_norm": 0.24069858398639982, | |
| "learning_rate": 1.6538298366257975e-07, | |
| "loss": 0.1691, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.23664302787938343, | |
| "learning_rate": 1.3902904145653094e-07, | |
| "loss": 0.1088, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 6.696969696969697, | |
| "grad_norm": 0.22795997559691616, | |
| "learning_rate": 1.1494637779369766e-07, | |
| "loss": 0.1222, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.2587396436241991, | |
| "learning_rate": 9.314053963669245e-08, | |
| "loss": 0.1401, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 6.757575757575758, | |
| "grad_norm": 0.24502445378712756, | |
| "learning_rate": 7.361654952665608e-08, | |
| "loss": 0.1748, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 6.787878787878788, | |
| "grad_norm": 0.22880608428050075, | |
| "learning_rate": 5.637890442641403e-08, | |
| "loss": 0.1418, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 6.818181818181818, | |
| "grad_norm": 0.2407871573216366, | |
| "learning_rate": 4.143157468468717e-08, | |
| "loss": 0.1669, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 6.848484848484849, | |
| "grad_norm": 0.24810977258480002, | |
| "learning_rate": 2.8778003121607834e-08, | |
| "loss": 0.175, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 6.878787878787879, | |
| "grad_norm": 0.22507015078779058, | |
| "learning_rate": 1.8421104235727406e-08, | |
| "loss": 0.1409, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.24240398869550214, | |
| "learning_rate": 1.0363263532724433e-08, | |
| "loss": 0.1763, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 6.9393939393939394, | |
| "grad_norm": 0.23820032801788155, | |
| "learning_rate": 4.606336975948589e-09, | |
| "loss": 0.1604, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 6.96969696969697, | |
| "grad_norm": 0.23457485569594724, | |
| "learning_rate": 1.1516505589381777e-09, | |
| "loss": 0.1425, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.24910710827382176, | |
| "learning_rate": 0.0, | |
| "loss": 0.138, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 231, | |
| "total_flos": 2.484313596666839e+17, | |
| "train_loss": 0.31677277253948766, | |
| "train_runtime": 4066.0104, | |
| "train_samples_per_second": 5.44, | |
| "train_steps_per_second": 0.057 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 231, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.484313596666839e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |