| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9979879275653922, | |
| "eval_steps": 47, | |
| "global_step": 558, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0053655264922870555, | |
| "grad_norm": 1.2779298491774165, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 2.0226, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0053655264922870555, | |
| "eval_loss": 3.5763909816741943, | |
| "eval_runtime": 403.4915, | |
| "eval_samples_per_second": 6.317, | |
| "eval_steps_per_second": 0.791, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.010731052984574111, | |
| "grad_norm": 0.9868395151451838, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 2.0251, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01609657947686117, | |
| "grad_norm": 1.124719320101948, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 2.0676, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.021462105969148222, | |
| "grad_norm": 0.8967993867216901, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 2.0685, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02682763246143528, | |
| "grad_norm": 0.8530329942799596, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 1.975, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03219315895372234, | |
| "grad_norm": 17.572932401470002, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 2.0319, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03755868544600939, | |
| "grad_norm": 1.5955614660997672, | |
| "learning_rate": 1.2962962962962962e-05, | |
| "loss": 1.9776, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.042924211938296444, | |
| "grad_norm": 1.0305843114650708, | |
| "learning_rate": 1.4814814814814815e-05, | |
| "loss": 1.9384, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0482897384305835, | |
| "grad_norm": 0.7931768664816836, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 2.0357, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05365526492287056, | |
| "grad_norm": 0.7458181490503686, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 1.9343, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05902079141515761, | |
| "grad_norm": 1.0379874858551188, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 1.9069, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06438631790744467, | |
| "grad_norm": 0.9634098196921086, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 1.8831, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.06975184439973173, | |
| "grad_norm": 0.8492217551097285, | |
| "learning_rate": 2.4074074074074074e-05, | |
| "loss": 1.9665, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07511737089201878, | |
| "grad_norm": 0.8866414055564286, | |
| "learning_rate": 2.5925925925925925e-05, | |
| "loss": 1.9859, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08048289738430583, | |
| "grad_norm": 0.8975616587977772, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 1.9621, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08584842387659289, | |
| "grad_norm": 0.8856060845121474, | |
| "learning_rate": 2.962962962962963e-05, | |
| "loss": 2.0504, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09121395036887994, | |
| "grad_norm": 0.7934430921825818, | |
| "learning_rate": 3.148148148148148e-05, | |
| "loss": 1.8965, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.096579476861167, | |
| "grad_norm": 0.8411958882279533, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 1.834, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10194500335345406, | |
| "grad_norm": 0.9605732267993595, | |
| "learning_rate": 3.518518518518519e-05, | |
| "loss": 2.0321, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.10731052984574112, | |
| "grad_norm": 0.8964432934098547, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 1.9165, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11267605633802817, | |
| "grad_norm": 0.8325547065304583, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 2.0088, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.11804158283031523, | |
| "grad_norm": 0.7337982135739225, | |
| "learning_rate": 4.074074074074074e-05, | |
| "loss": 1.8652, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.12340710932260228, | |
| "grad_norm": 0.8703749413838572, | |
| "learning_rate": 4.259259259259259e-05, | |
| "loss": 1.9384, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.12877263581488935, | |
| "grad_norm": 0.752681510079145, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 1.84, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1341381623071764, | |
| "grad_norm": 0.7664305287185245, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 1.917, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.13950368879946345, | |
| "grad_norm": 0.7376417365374465, | |
| "learning_rate": 4.814814814814815e-05, | |
| "loss": 1.8995, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1448692152917505, | |
| "grad_norm": 0.6723119101020634, | |
| "learning_rate": 5e-05, | |
| "loss": 1.8572, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15023474178403756, | |
| "grad_norm": 0.6636335261172684, | |
| "learning_rate": 4.999956245830044e-05, | |
| "loss": 1.8311, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.15560026827632462, | |
| "grad_norm": 0.7118137563548644, | |
| "learning_rate": 4.9998249848517185e-05, | |
| "loss": 1.9461, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.16096579476861167, | |
| "grad_norm": 0.5571924001587539, | |
| "learning_rate": 4.999606221659595e-05, | |
| "loss": 1.9125, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.16633132126089872, | |
| "grad_norm": 0.5895748551871794, | |
| "learning_rate": 4.999299963911115e-05, | |
| "loss": 1.9295, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.17169684775318578, | |
| "grad_norm": 0.5114717351226575, | |
| "learning_rate": 4.9989062223263216e-05, | |
| "loss": 1.8642, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.17706237424547283, | |
| "grad_norm": 0.5658751591114588, | |
| "learning_rate": 4.998425010687484e-05, | |
| "loss": 1.9465, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.18242790073775988, | |
| "grad_norm": 0.4198888613535172, | |
| "learning_rate": 4.997856345838615e-05, | |
| "loss": 1.7824, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.18779342723004694, | |
| "grad_norm": 0.558439585084249, | |
| "learning_rate": 4.99720024768488e-05, | |
| "loss": 1.9157, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.193158953722334, | |
| "grad_norm": 0.4888963940737741, | |
| "learning_rate": 4.996456739191905e-05, | |
| "loss": 1.9136, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.19852448021462105, | |
| "grad_norm": 0.5106753170932593, | |
| "learning_rate": 4.995625846384966e-05, | |
| "loss": 1.8623, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.20389000670690813, | |
| "grad_norm": 0.5942581620888907, | |
| "learning_rate": 4.994707598348085e-05, | |
| "loss": 1.8741, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.20925553319919518, | |
| "grad_norm": 0.5106367767120957, | |
| "learning_rate": 4.993702027223004e-05, | |
| "loss": 1.9413, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.21462105969148224, | |
| "grad_norm": 0.5113103948113709, | |
| "learning_rate": 4.992609168208069e-05, | |
| "loss": 1.9966, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2199865861837693, | |
| "grad_norm": 0.5212461839914251, | |
| "learning_rate": 4.9914290595569895e-05, | |
| "loss": 1.8474, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.22535211267605634, | |
| "grad_norm": 0.5161668483087154, | |
| "learning_rate": 4.9901617425775067e-05, | |
| "loss": 1.8557, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2307176391683434, | |
| "grad_norm": 0.5484719777790631, | |
| "learning_rate": 4.988807261629942e-05, | |
| "loss": 1.8769, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.23608316566063045, | |
| "grad_norm": 0.4937411819105057, | |
| "learning_rate": 4.987365664125647e-05, | |
| "loss": 1.8923, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2414486921529175, | |
| "grad_norm": 0.5804971756745295, | |
| "learning_rate": 4.985837000525343e-05, | |
| "loss": 1.954, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.24681421864520456, | |
| "grad_norm": 0.531027559174979, | |
| "learning_rate": 4.984221324337356e-05, | |
| "loss": 1.9758, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.25217974513749164, | |
| "grad_norm": 0.5132447531432143, | |
| "learning_rate": 4.982518692115744e-05, | |
| "loss": 1.8588, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.25217974513749164, | |
| "eval_loss": 2.2616188526153564, | |
| "eval_runtime": 403.9965, | |
| "eval_samples_per_second": 6.309, | |
| "eval_steps_per_second": 0.79, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2575452716297787, | |
| "grad_norm": 0.4799971353542736, | |
| "learning_rate": 4.980729163458312e-05, | |
| "loss": 1.8558, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.26291079812206575, | |
| "grad_norm": 0.4318157242619782, | |
| "learning_rate": 4.978852801004534e-05, | |
| "loss": 1.9574, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2682763246143528, | |
| "grad_norm": 0.44416051699732234, | |
| "learning_rate": 4.976889670433355e-05, | |
| "loss": 1.7988, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.27364185110663986, | |
| "grad_norm": 0.42380891285410505, | |
| "learning_rate": 4.974839840460895e-05, | |
| "loss": 1.8296, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2790073775989269, | |
| "grad_norm": 0.4989220387491634, | |
| "learning_rate": 4.97270338283804e-05, | |
| "loss": 1.8996, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.28437290409121396, | |
| "grad_norm": 0.44554710351703636, | |
| "learning_rate": 4.970480372347934e-05, | |
| "loss": 1.9553, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.289738430583501, | |
| "grad_norm": 0.5597504355268026, | |
| "learning_rate": 4.9681708868033616e-05, | |
| "loss": 1.9979, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.29510395707578807, | |
| "grad_norm": 0.4758356280331154, | |
| "learning_rate": 4.9657750070440196e-05, | |
| "loss": 1.9244, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3004694835680751, | |
| "grad_norm": 0.4337532785073617, | |
| "learning_rate": 4.963292816933692e-05, | |
| "loss": 1.9099, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3058350100603622, | |
| "grad_norm": 0.48193864337289033, | |
| "learning_rate": 4.9607244033573156e-05, | |
| "loss": 1.8076, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.31120053655264923, | |
| "grad_norm": 0.5180685237569209, | |
| "learning_rate": 4.9580698562179297e-05, | |
| "loss": 1.9031, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.3165660630449363, | |
| "grad_norm": 0.4227856708321325, | |
| "learning_rate": 4.955329268433543e-05, | |
| "loss": 1.8111, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.32193158953722334, | |
| "grad_norm": 0.4457771321490526, | |
| "learning_rate": 4.9525027359338696e-05, | |
| "loss": 1.8771, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3272971160295104, | |
| "grad_norm": 0.383809821251567, | |
| "learning_rate": 4.949590357656975e-05, | |
| "loss": 1.8384, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.33266264252179745, | |
| "grad_norm": 0.5245332667935808, | |
| "learning_rate": 4.946592235545815e-05, | |
| "loss": 1.948, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.3380281690140845, | |
| "grad_norm": 0.48131587989680386, | |
| "learning_rate": 4.9435084745446666e-05, | |
| "loss": 1.9546, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.34339369550637155, | |
| "grad_norm": 0.47579745445020977, | |
| "learning_rate": 4.940339182595451e-05, | |
| "loss": 1.9085, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.3487592219986586, | |
| "grad_norm": 0.5162389548444078, | |
| "learning_rate": 4.9370844706339594e-05, | |
| "loss": 1.9263, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.35412474849094566, | |
| "grad_norm": 0.4430593262438425, | |
| "learning_rate": 4.933744452585966e-05, | |
| "loss": 1.8264, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.3594902749832327, | |
| "grad_norm": 0.4564426740705147, | |
| "learning_rate": 4.930319245363248e-05, | |
| "loss": 1.7974, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.36485580147551977, | |
| "grad_norm": 0.4896650401826788, | |
| "learning_rate": 4.926808968859483e-05, | |
| "loss": 1.8567, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3702213279678068, | |
| "grad_norm": 0.4051071793789495, | |
| "learning_rate": 4.923213745946059e-05, | |
| "loss": 1.8201, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3755868544600939, | |
| "grad_norm": 0.48978661474326063, | |
| "learning_rate": 4.919533702467771e-05, | |
| "loss": 1.8006, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 0.5175966382748892, | |
| "learning_rate": 4.9157689672384174e-05, | |
| "loss": 1.7904, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.386317907444668, | |
| "grad_norm": 0.49081107125596507, | |
| "learning_rate": 4.91191967203629e-05, | |
| "loss": 1.9606, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.39168343393695504, | |
| "grad_norm": 0.5920859101699153, | |
| "learning_rate": 4.907985951599563e-05, | |
| "loss": 1.8124, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.3970489604292421, | |
| "grad_norm": 0.4106853545996278, | |
| "learning_rate": 4.9039679436215734e-05, | |
| "loss": 1.9402, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4024144869215292, | |
| "grad_norm": 0.6429573337250454, | |
| "learning_rate": 4.899865788746005e-05, | |
| "loss": 1.8233, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.40778001341381626, | |
| "grad_norm": 0.5770635817407002, | |
| "learning_rate": 4.895679630561963e-05, | |
| "loss": 1.8909, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.4131455399061033, | |
| "grad_norm": 0.8202937145800111, | |
| "learning_rate": 4.891409615598949e-05, | |
| "loss": 1.8635, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.41851106639839036, | |
| "grad_norm": 0.6871160286583309, | |
| "learning_rate": 4.88705589332173e-05, | |
| "loss": 1.9497, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4238765928906774, | |
| "grad_norm": 0.4438407032445468, | |
| "learning_rate": 4.882618616125111e-05, | |
| "loss": 1.9336, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.42924211938296447, | |
| "grad_norm": 0.545225076259868, | |
| "learning_rate": 4.878097939328596e-05, | |
| "loss": 1.843, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4346076458752515, | |
| "grad_norm": 0.5120969176703908, | |
| "learning_rate": 4.873494021170953e-05, | |
| "loss": 1.9019, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.4399731723675386, | |
| "grad_norm": 0.41579446347477983, | |
| "learning_rate": 4.868807022804678e-05, | |
| "loss": 1.8617, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.44533869885982563, | |
| "grad_norm": 0.5510693357302907, | |
| "learning_rate": 4.864037108290347e-05, | |
| "loss": 1.9433, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4507042253521127, | |
| "grad_norm": 0.40913367019675556, | |
| "learning_rate": 4.859184444590882e-05, | |
| "loss": 1.8353, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.45606975184439974, | |
| "grad_norm": 0.3763529830498608, | |
| "learning_rate": 4.854249201565701e-05, | |
| "loss": 1.7299, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4614352783366868, | |
| "grad_norm": 0.47448012995411376, | |
| "learning_rate": 4.849231551964771e-05, | |
| "loss": 1.8547, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.46680080482897385, | |
| "grad_norm": 0.4181797287844973, | |
| "learning_rate": 4.84413167142257e-05, | |
| "loss": 1.8186, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.4721663313212609, | |
| "grad_norm": 0.4781785535505977, | |
| "learning_rate": 4.838949738451929e-05, | |
| "loss": 1.8965, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.47753185781354796, | |
| "grad_norm": 0.4351347675471463, | |
| "learning_rate": 4.833685934437787e-05, | |
| "loss": 1.863, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.482897384305835, | |
| "grad_norm": 0.3874884362959953, | |
| "learning_rate": 4.8283404436308464e-05, | |
| "loss": 1.7396, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.48826291079812206, | |
| "grad_norm": 0.49444219921157534, | |
| "learning_rate": 4.8229134531411166e-05, | |
| "loss": 1.8444, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4936284372904091, | |
| "grad_norm": 0.39922493357750344, | |
| "learning_rate": 4.8174051529313704e-05, | |
| "loss": 1.8247, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.49899396378269617, | |
| "grad_norm": 0.5511215543224721, | |
| "learning_rate": 4.81181573581049e-05, | |
| "loss": 1.9533, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5043594902749833, | |
| "grad_norm": 0.41108793496635143, | |
| "learning_rate": 4.8061453974267195e-05, | |
| "loss": 1.7952, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5043594902749833, | |
| "eval_loss": 2.198935031890869, | |
| "eval_runtime": 403.21, | |
| "eval_samples_per_second": 6.322, | |
| "eval_steps_per_second": 0.791, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5097250167672703, | |
| "grad_norm": 0.43234228757508586, | |
| "learning_rate": 4.80039433626082e-05, | |
| "loss": 1.8667, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5150905432595574, | |
| "grad_norm": 0.40707489762011484, | |
| "learning_rate": 4.7945627536191166e-05, | |
| "loss": 1.9081, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5204560697518444, | |
| "grad_norm": 0.41662703960667113, | |
| "learning_rate": 4.788650853626456e-05, | |
| "loss": 1.7536, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5258215962441315, | |
| "grad_norm": 0.39754454922458654, | |
| "learning_rate": 4.7826588432190614e-05, | |
| "loss": 1.8851, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5311871227364185, | |
| "grad_norm": 0.41218018944985396, | |
| "learning_rate": 4.7765869321372836e-05, | |
| "loss": 1.8062, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5365526492287056, | |
| "grad_norm": 0.41546782428712065, | |
| "learning_rate": 4.7704353329182673e-05, | |
| "loss": 1.8491, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5419181757209927, | |
| "grad_norm": 0.47175537335479406, | |
| "learning_rate": 4.7642042608885064e-05, | |
| "loss": 1.7917, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5472837022132797, | |
| "grad_norm": 0.4373071645022652, | |
| "learning_rate": 4.7578939341563095e-05, | |
| "loss": 1.8632, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5526492287055668, | |
| "grad_norm": 0.4369286574512984, | |
| "learning_rate": 4.751504573604162e-05, | |
| "loss": 1.8377, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5580147551978538, | |
| "grad_norm": 0.40840919093725575, | |
| "learning_rate": 4.745036402880999e-05, | |
| "loss": 1.8397, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5633802816901409, | |
| "grad_norm": 0.4023593789216688, | |
| "learning_rate": 4.738489648394373e-05, | |
| "loss": 1.9389, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5687458081824279, | |
| "grad_norm": 0.45139851873493203, | |
| "learning_rate": 4.731864539302531e-05, | |
| "loss": 1.9384, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.574111334674715, | |
| "grad_norm": 0.518973698127367, | |
| "learning_rate": 4.725161307506391e-05, | |
| "loss": 1.8388, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.579476861167002, | |
| "grad_norm": 0.41440109186821383, | |
| "learning_rate": 4.7183801876414294e-05, | |
| "loss": 1.9807, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5848423876592891, | |
| "grad_norm": 0.485066583270121, | |
| "learning_rate": 4.711521417069462e-05, | |
| "loss": 1.9518, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5902079141515761, | |
| "grad_norm": 0.453283667200083, | |
| "learning_rate": 4.70458523587034e-05, | |
| "loss": 1.8874, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5955734406438632, | |
| "grad_norm": 0.40638037542695366, | |
| "learning_rate": 4.697571886833544e-05, | |
| "loss": 1.8138, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6009389671361502, | |
| "grad_norm": 0.47718260435002885, | |
| "learning_rate": 4.6904816154496854e-05, | |
| "loss": 1.8409, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6063044936284373, | |
| "grad_norm": 0.45172842172725275, | |
| "learning_rate": 4.683314669901918e-05, | |
| "loss": 1.9261, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6116700201207244, | |
| "grad_norm": 0.4713867113135493, | |
| "learning_rate": 4.676071301057243e-05, | |
| "loss": 1.8502, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6170355466130114, | |
| "grad_norm": 0.42451831321750794, | |
| "learning_rate": 4.668751762457734e-05, | |
| "loss": 1.8489, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6224010731052985, | |
| "grad_norm": 0.3958564758105134, | |
| "learning_rate": 4.6613563103116594e-05, | |
| "loss": 1.8922, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6277665995975855, | |
| "grad_norm": 0.4009292301802374, | |
| "learning_rate": 4.653885203484515e-05, | |
| "loss": 1.8685, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6331321260898726, | |
| "grad_norm": 0.37598195515170946, | |
| "learning_rate": 4.6463387034899645e-05, | |
| "loss": 1.7889, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6384976525821596, | |
| "grad_norm": 0.4710597532282601, | |
| "learning_rate": 4.638717074480682e-05, | |
| "loss": 1.8739, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6438631790744467, | |
| "grad_norm": 0.4146663267990981, | |
| "learning_rate": 4.631020583239107e-05, | |
| "loss": 1.8988, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6492287055667337, | |
| "grad_norm": 0.4354294422619394, | |
| "learning_rate": 4.6232494991681094e-05, | |
| "loss": 1.8517, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6545942320590208, | |
| "grad_norm": 0.5470562094587934, | |
| "learning_rate": 4.615404094281554e-05, | |
| "loss": 1.812, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6599597585513078, | |
| "grad_norm": 0.37147901827060426, | |
| "learning_rate": 4.607484643194788e-05, | |
| "loss": 1.8447, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6653252850435949, | |
| "grad_norm": 0.462225949803499, | |
| "learning_rate": 4.599491423115014e-05, | |
| "loss": 1.8442, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.670690811535882, | |
| "grad_norm": 0.4600536423365438, | |
| "learning_rate": 4.5914247138316025e-05, | |
| "loss": 1.926, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.676056338028169, | |
| "grad_norm": 0.4049920788749757, | |
| "learning_rate": 4.5832847977062874e-05, | |
| "loss": 1.8001, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.681421864520456, | |
| "grad_norm": 0.4052262687972146, | |
| "learning_rate": 4.5750719596632885e-05, | |
| "loss": 1.8616, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6867873910127431, | |
| "grad_norm": 0.455611587260568, | |
| "learning_rate": 4.5667864871793345e-05, | |
| "loss": 1.7792, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6921529175050302, | |
| "grad_norm": 0.4321930788528261, | |
| "learning_rate": 4.558428670273601e-05, | |
| "loss": 1.911, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.6975184439973172, | |
| "grad_norm": 0.3922798652619637, | |
| "learning_rate": 4.549998801497564e-05, | |
| "loss": 1.8979, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7028839704896043, | |
| "grad_norm": 0.4338149926251678, | |
| "learning_rate": 4.54149717592475e-05, | |
| "loss": 1.793, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7082494969818913, | |
| "grad_norm": 0.42943405977885124, | |
| "learning_rate": 4.532924091140417e-05, | |
| "loss": 1.8606, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7136150234741784, | |
| "grad_norm": 0.4145321079739444, | |
| "learning_rate": 4.524279847231131e-05, | |
| "loss": 1.906, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7189805499664654, | |
| "grad_norm": 0.38959038753607894, | |
| "learning_rate": 4.515564746774265e-05, | |
| "loss": 1.7323, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7243460764587525, | |
| "grad_norm": 0.49146581928185773, | |
| "learning_rate": 4.5067790948274094e-05, | |
| "loss": 1.8303, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7297116029510395, | |
| "grad_norm": 0.4638570517033698, | |
| "learning_rate": 4.49792319891769e-05, | |
| "loss": 1.8243, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7350771294433266, | |
| "grad_norm": 0.38160033679573113, | |
| "learning_rate": 4.4889973690310085e-05, | |
| "loss": 1.8186, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7404426559356136, | |
| "grad_norm": 0.5253837731077495, | |
| "learning_rate": 4.480001917601185e-05, | |
| "loss": 1.9086, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7458081824279007, | |
| "grad_norm": 0.5010255463058967, | |
| "learning_rate": 4.470937159499029e-05, | |
| "loss": 1.9014, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7511737089201878, | |
| "grad_norm": 0.44408252888412136, | |
| "learning_rate": 4.461803412021314e-05, | |
| "loss": 1.886, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7565392354124748, | |
| "grad_norm": 0.5189286834612553, | |
| "learning_rate": 4.4526009948796703e-05, | |
| "loss": 1.8266, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7565392354124748, | |
| "eval_loss": 2.1618716716766357, | |
| "eval_runtime": 403.938, | |
| "eval_samples_per_second": 6.31, | |
| "eval_steps_per_second": 0.79, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 0.4334524920814895, | |
| "learning_rate": 4.4433302301893987e-05, | |
| "loss": 1.8863, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7672702883970489, | |
| "grad_norm": 0.4809144135248221, | |
| "learning_rate": 4.433991442458188e-05, | |
| "loss": 1.9075, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.772635814889336, | |
| "grad_norm": 0.5258910256081301, | |
| "learning_rate": 4.4245849585747654e-05, | |
| "loss": 1.9616, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.778001341381623, | |
| "grad_norm": 0.5131536872616902, | |
| "learning_rate": 4.415111107797445e-05, | |
| "loss": 1.8953, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7833668678739101, | |
| "grad_norm": 0.45216150555891305, | |
| "learning_rate": 4.4055702217426084e-05, | |
| "loss": 1.8867, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.7887323943661971, | |
| "grad_norm": 0.47907057676705184, | |
| "learning_rate": 4.395962634373097e-05, | |
| "loss": 1.8335, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.7940979208584842, | |
| "grad_norm": 0.4945502094720473, | |
| "learning_rate": 4.386288681986516e-05, | |
| "loss": 1.9218, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.7994634473507712, | |
| "grad_norm": 0.4595838282230734, | |
| "learning_rate": 4.376548703203474e-05, | |
| "loss": 1.8603, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8048289738430584, | |
| "grad_norm": 0.45448249598162854, | |
| "learning_rate": 4.36674303895572e-05, | |
| "loss": 1.8615, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8101945003353455, | |
| "grad_norm": 0.4238860646236377, | |
| "learning_rate": 4.356872032474213e-05, | |
| "loss": 1.8374, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8155600268276325, | |
| "grad_norm": 0.4519762198634726, | |
| "learning_rate": 4.34693602927711e-05, | |
| "loss": 1.8174, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8209255533199196, | |
| "grad_norm": 0.45624612554196736, | |
| "learning_rate": 4.336935377157668e-05, | |
| "loss": 1.9105, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8262910798122066, | |
| "grad_norm": 0.4076679531263349, | |
| "learning_rate": 4.326870426172075e-05, | |
| "loss": 1.8865, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8316566063044937, | |
| "grad_norm": 0.4514412896189259, | |
| "learning_rate": 4.3167415286271905e-05, | |
| "loss": 1.9272, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8370221327967807, | |
| "grad_norm": 0.40433175857719245, | |
| "learning_rate": 4.3065490390682186e-05, | |
| "loss": 1.814, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8423876592890678, | |
| "grad_norm": 0.4159285646615437, | |
| "learning_rate": 4.296293314266294e-05, | |
| "loss": 1.8064, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.8477531857813548, | |
| "grad_norm": 0.4270220599999327, | |
| "learning_rate": 4.2859747132060006e-05, | |
| "loss": 1.7482, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.8531187122736419, | |
| "grad_norm": 0.5162993182758251, | |
| "learning_rate": 4.275593597072796e-05, | |
| "loss": 1.7763, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.8584842387659289, | |
| "grad_norm": 0.4065368416357515, | |
| "learning_rate": 4.265150329240376e-05, | |
| "loss": 1.885, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.863849765258216, | |
| "grad_norm": 0.5523980888304064, | |
| "learning_rate": 4.2546452752579536e-05, | |
| "loss": 1.971, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.869215291750503, | |
| "grad_norm": 0.4121779650999884, | |
| "learning_rate": 4.2440788028374624e-05, | |
| "loss": 1.9301, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.8745808182427901, | |
| "grad_norm": 0.4612759438023989, | |
| "learning_rate": 4.233451281840686e-05, | |
| "loss": 1.8564, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.8799463447350772, | |
| "grad_norm": 0.3977527412707747, | |
| "learning_rate": 4.2227630842663136e-05, | |
| "loss": 1.8876, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.8853118712273642, | |
| "grad_norm": 0.45276964005664955, | |
| "learning_rate": 4.212014584236914e-05, | |
| "loss": 1.8098, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8906773977196513, | |
| "grad_norm": 0.37349476483277844, | |
| "learning_rate": 4.2012061579858465e-05, | |
| "loss": 1.8247, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.8960429242119383, | |
| "grad_norm": 0.39676844111499676, | |
| "learning_rate": 4.190338183844086e-05, | |
| "loss": 1.914, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9014084507042254, | |
| "grad_norm": 0.3625691966854791, | |
| "learning_rate": 4.1794110422269825e-05, | |
| "loss": 1.92, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9067739771965124, | |
| "grad_norm": 0.39688352111910685, | |
| "learning_rate": 4.168425115620944e-05, | |
| "loss": 1.8103, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9121395036887995, | |
| "grad_norm": 0.3737222783168723, | |
| "learning_rate": 4.157380788570053e-05, | |
| "loss": 1.8215, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9175050301810865, | |
| "grad_norm": 0.43842445138079555, | |
| "learning_rate": 4.146278447662597e-05, | |
| "loss": 1.8029, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9228705566733736, | |
| "grad_norm": 0.47414715687042996, | |
| "learning_rate": 4.1351184815175456e-05, | |
| "loss": 1.8974, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9282360831656606, | |
| "grad_norm": 0.4146539941665616, | |
| "learning_rate": 4.123901280770945e-05, | |
| "loss": 1.8871, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9336016096579477, | |
| "grad_norm": 0.4088656394439229, | |
| "learning_rate": 4.112627238062239e-05, | |
| "loss": 1.8594, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9389671361502347, | |
| "grad_norm": 0.40239314445747604, | |
| "learning_rate": 4.101296748020533e-05, | |
| "loss": 1.8207, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9443326626425218, | |
| "grad_norm": 0.36389736696587754, | |
| "learning_rate": 4.089910207250778e-05, | |
| "loss": 1.8423, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.9496981891348089, | |
| "grad_norm": 0.41161031740209597, | |
| "learning_rate": 4.0784680143198836e-05, | |
| "loss": 1.8954, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.9550637156270959, | |
| "grad_norm": 0.3613771597005119, | |
| "learning_rate": 4.0669705697427754e-05, | |
| "loss": 1.9036, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.960429242119383, | |
| "grad_norm": 0.36880689305175635, | |
| "learning_rate": 4.055418275968368e-05, | |
| "loss": 1.8676, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.96579476861167, | |
| "grad_norm": 0.3892009514188599, | |
| "learning_rate": 4.04381153736548e-05, | |
| "loss": 1.8545, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9711602951039571, | |
| "grad_norm": 0.39386413560247385, | |
| "learning_rate": 4.032150760208684e-05, | |
| "loss": 1.9306, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.9765258215962441, | |
| "grad_norm": 0.38085985815329654, | |
| "learning_rate": 4.02043635266408e-05, | |
| "loss": 1.9339, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.9818913480885312, | |
| "grad_norm": 0.3502551322824309, | |
| "learning_rate": 4.00866872477501e-05, | |
| "loss": 1.8438, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.9872568745808182, | |
| "grad_norm": 0.42923155491093273, | |
| "learning_rate": 3.9968482884477075e-05, | |
| "loss": 1.8326, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.9926224010731053, | |
| "grad_norm": 0.37629854526288287, | |
| "learning_rate": 3.9849754574368766e-05, | |
| "loss": 1.8521, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.9979879275653923, | |
| "grad_norm": 0.43648587047407056, | |
| "learning_rate": 3.973050647331209e-05, | |
| "loss": 1.9192, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.0053655264922872, | |
| "grad_norm": 1.1037688215019685, | |
| "learning_rate": 3.9610742755388406e-05, | |
| "loss": 3.5179, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.010731052984574, | |
| "grad_norm": 0.5315167109100908, | |
| "learning_rate": 3.949046761272736e-05, | |
| "loss": 1.5997, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.010731052984574, | |
| "eval_loss": 2.160255193710327, | |
| "eval_runtime": 403.1754, | |
| "eval_samples_per_second": 6.322, | |
| "eval_steps_per_second": 0.791, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.0160965794768613, | |
| "grad_norm": 1.034163349005432, | |
| "learning_rate": 3.9369685255360175e-05, | |
| "loss": 1.6376, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.0214621059691482, | |
| "grad_norm": 0.5297370695526916, | |
| "learning_rate": 3.924839991107229e-05, | |
| "loss": 1.5196, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0268276324614354, | |
| "grad_norm": 0.49978902813452, | |
| "learning_rate": 3.9126615825255364e-05, | |
| "loss": 1.4628, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.0321931589537223, | |
| "grad_norm": 0.5207450843338, | |
| "learning_rate": 3.900433726075865e-05, | |
| "loss": 1.5808, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.0375586854460095, | |
| "grad_norm": 0.5502855697944308, | |
| "learning_rate": 3.888156849773985e-05, | |
| "loss": 1.5445, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.0429242119382964, | |
| "grad_norm": 0.5175112784168553, | |
| "learning_rate": 3.875831383351519e-05, | |
| "loss": 1.6456, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.0482897384305836, | |
| "grad_norm": 0.4649959629509902, | |
| "learning_rate": 3.863457758240912e-05, | |
| "loss": 1.5508, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.0536552649228705, | |
| "grad_norm": 0.48136239187851326, | |
| "learning_rate": 3.851036407560319e-05, | |
| "loss": 1.4848, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.0590207914151577, | |
| "grad_norm": 0.48389312973032594, | |
| "learning_rate": 3.838567766098452e-05, | |
| "loss": 1.5123, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.0643863179074446, | |
| "grad_norm": 0.45180450785183063, | |
| "learning_rate": 3.826052270299356e-05, | |
| "loss": 1.5807, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.0697518443997318, | |
| "grad_norm": 0.48079670948977926, | |
| "learning_rate": 3.813490358247137e-05, | |
| "loss": 1.5426, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.0751173708920188, | |
| "grad_norm": 0.510291013526601, | |
| "learning_rate": 3.800882469650621e-05, | |
| "loss": 1.5845, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.080482897384306, | |
| "grad_norm": 0.41649615075440893, | |
| "learning_rate": 3.78822904582797e-05, | |
| "loss": 1.5003, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.0858484238765929, | |
| "grad_norm": 0.4499681852447958, | |
| "learning_rate": 3.7755305296912276e-05, | |
| "loss": 1.4751, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.09121395036888, | |
| "grad_norm": 0.5088284534955508, | |
| "learning_rate": 3.762787365730821e-05, | |
| "loss": 1.5982, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.096579476861167, | |
| "grad_norm": 0.3929831451008196, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.4991, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.1019450033534541, | |
| "grad_norm": 0.41107315409379536, | |
| "learning_rate": 3.7371688800992235e-05, | |
| "loss": 1.4837, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.107310529845741, | |
| "grad_norm": 0.415588978346341, | |
| "learning_rate": 3.7242944551604914e-05, | |
| "loss": 1.5345, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.1126760563380282, | |
| "grad_norm": 0.4044749735529877, | |
| "learning_rate": 3.711377175831626e-05, | |
| "loss": 1.4545, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.1180415828303152, | |
| "grad_norm": 0.44865460732176465, | |
| "learning_rate": 3.698417494260494e-05, | |
| "loss": 1.6391, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.1234071093226023, | |
| "grad_norm": 0.4479424002899852, | |
| "learning_rate": 3.685415864079185e-05, | |
| "loss": 1.508, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.1287726358148893, | |
| "grad_norm": 0.429927217929734, | |
| "learning_rate": 3.6723727403881284e-05, | |
| "loss": 1.5501, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1341381623071765, | |
| "grad_norm": 0.5044115066105113, | |
| "learning_rate": 3.659288579740163e-05, | |
| "loss": 1.5615, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.1395036887994634, | |
| "grad_norm": 0.424575600191598, | |
| "learning_rate": 3.646163840124561e-05, | |
| "loss": 1.4835, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.1448692152917506, | |
| "grad_norm": 0.43480586168977786, | |
| "learning_rate": 3.632998980950993e-05, | |
| "loss": 1.5371, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.1502347417840375, | |
| "grad_norm": 0.40322150224409975, | |
| "learning_rate": 3.619794463033447e-05, | |
| "loss": 1.5376, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.1556002682763247, | |
| "grad_norm": 0.42637706485283583, | |
| "learning_rate": 3.6065507485741e-05, | |
| "loss": 1.5631, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.1609657947686116, | |
| "grad_norm": 0.4259143981435151, | |
| "learning_rate": 3.593268301147139e-05, | |
| "loss": 1.5741, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.1663313212608988, | |
| "grad_norm": 0.45826011738708783, | |
| "learning_rate": 3.5799475856825326e-05, | |
| "loss": 1.5298, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.1716968477531857, | |
| "grad_norm": 0.4030875431169561, | |
| "learning_rate": 3.566589068449761e-05, | |
| "loss": 1.5574, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.1770623742454729, | |
| "grad_norm": 0.409604426981432, | |
| "learning_rate": 3.5531932170414896e-05, | |
| "loss": 1.4859, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.1824279007377598, | |
| "grad_norm": 0.37439697455127857, | |
| "learning_rate": 3.539760500357207e-05, | |
| "loss": 1.5221, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.187793427230047, | |
| "grad_norm": 0.3600455315943667, | |
| "learning_rate": 3.5262913885868066e-05, | |
| "loss": 1.479, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.193158953722334, | |
| "grad_norm": 0.39515964344444054, | |
| "learning_rate": 3.512786353194134e-05, | |
| "loss": 1.6002, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.198524480214621, | |
| "grad_norm": 0.3903215407425732, | |
| "learning_rate": 3.49924586690048e-05, | |
| "loss": 1.5627, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.203890006706908, | |
| "grad_norm": 0.3863928913911809, | |
| "learning_rate": 3.485670403668036e-05, | |
| "loss": 1.4894, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.2092555331991952, | |
| "grad_norm": 0.39120322752660747, | |
| "learning_rate": 3.472060438683302e-05, | |
| "loss": 1.5576, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.2146210596914822, | |
| "grad_norm": 0.4012276022035902, | |
| "learning_rate": 3.4584164483404544e-05, | |
| "loss": 1.5661, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.2199865861837693, | |
| "grad_norm": 0.359333036872216, | |
| "learning_rate": 3.444738910224671e-05, | |
| "loss": 1.6043, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.2253521126760563, | |
| "grad_norm": 0.4169926695855886, | |
| "learning_rate": 3.431028303095415e-05, | |
| "loss": 1.5162, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.2307176391683434, | |
| "grad_norm": 0.3666243644496504, | |
| "learning_rate": 3.417285106869673e-05, | |
| "loss": 1.5182, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.2360831656606304, | |
| "grad_norm": 0.3732640706992433, | |
| "learning_rate": 3.403509802605159e-05, | |
| "loss": 1.6136, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.2414486921529175, | |
| "grad_norm": 0.38533992909636827, | |
| "learning_rate": 3.389702872483477e-05, | |
| "loss": 1.5763, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.2468142186452045, | |
| "grad_norm": 0.3621197211070038, | |
| "learning_rate": 3.3758647997932417e-05, | |
| "loss": 1.5977, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.2521797451374916, | |
| "grad_norm": 0.39624430912627595, | |
| "learning_rate": 3.361996068913159e-05, | |
| "loss": 1.5794, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.2575452716297786, | |
| "grad_norm": 0.38404789088021773, | |
| "learning_rate": 3.348097165295076e-05, | |
| "loss": 1.5633, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.2629107981220657, | |
| "grad_norm": 0.35832957307552504, | |
| "learning_rate": 3.3341685754469856e-05, | |
| "loss": 1.5134, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.2629107981220657, | |
| "eval_loss": 2.15519642829895, | |
| "eval_runtime": 403.4218, | |
| "eval_samples_per_second": 6.318, | |
| "eval_steps_per_second": 0.791, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.268276324614353, | |
| "grad_norm": 0.34366404066845024, | |
| "learning_rate": 3.320210786915997e-05, | |
| "loss": 1.4825, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.2736418511066399, | |
| "grad_norm": 0.45636968737684863, | |
| "learning_rate": 3.3062242882712724e-05, | |
| "loss": 1.5146, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.2790073775989268, | |
| "grad_norm": 0.3804000236233779, | |
| "learning_rate": 3.2922095690869224e-05, | |
| "loss": 1.5501, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.284372904091214, | |
| "grad_norm": 0.3699794774550946, | |
| "learning_rate": 3.278167119924872e-05, | |
| "loss": 1.4956, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.2897384305835011, | |
| "grad_norm": 0.38134052511434596, | |
| "learning_rate": 3.2640974323176846e-05, | |
| "loss": 1.4926, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.295103957075788, | |
| "grad_norm": 0.4001577679575819, | |
| "learning_rate": 3.2500009987513655e-05, | |
| "loss": 1.5339, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.300469483568075, | |
| "grad_norm": 0.33599496371938614, | |
| "learning_rate": 3.235878312648112e-05, | |
| "loss": 1.3329, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.3058350100603622, | |
| "grad_norm": 0.3780832169945631, | |
| "learning_rate": 3.2217298683490525e-05, | |
| "loss": 1.5711, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.3112005365526493, | |
| "grad_norm": 0.37979912787014874, | |
| "learning_rate": 3.207556161096935e-05, | |
| "loss": 1.559, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.3165660630449363, | |
| "grad_norm": 0.34013678855872914, | |
| "learning_rate": 3.193357687018798e-05, | |
| "loss": 1.5112, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.3219315895372232, | |
| "grad_norm": 0.3761770494247479, | |
| "learning_rate": 3.179134943108597e-05, | |
| "loss": 1.5195, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.3272971160295104, | |
| "grad_norm": 0.36046502801340735, | |
| "learning_rate": 3.164888427209818e-05, | |
| "loss": 1.4648, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.3326626425217976, | |
| "grad_norm": 0.4163102856970103, | |
| "learning_rate": 3.150618637998041e-05, | |
| "loss": 1.6488, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.3380281690140845, | |
| "grad_norm": 0.37250576413738173, | |
| "learning_rate": 3.136326074963494e-05, | |
| "loss": 1.5966, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.3433936955063714, | |
| "grad_norm": 0.3717314773044194, | |
| "learning_rate": 3.122011238393562e-05, | |
| "loss": 1.4555, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3487592219986586, | |
| "grad_norm": 0.4233051858413458, | |
| "learning_rate": 3.1076746293552786e-05, | |
| "loss": 1.5931, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.3541247484909458, | |
| "grad_norm": 0.3538900779699424, | |
| "learning_rate": 3.093316749677788e-05, | |
| "loss": 1.5041, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.3594902749832327, | |
| "grad_norm": 0.4239252371757693, | |
| "learning_rate": 3.078938101934773e-05, | |
| "loss": 1.5986, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.3648558014755197, | |
| "grad_norm": 0.40455165331708126, | |
| "learning_rate": 3.064539189426874e-05, | |
| "loss": 1.5956, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.3702213279678068, | |
| "grad_norm": 0.35682161127333073, | |
| "learning_rate": 3.050120516164062e-05, | |
| "loss": 1.4983, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.375586854460094, | |
| "grad_norm": 0.366685178903664, | |
| "learning_rate": 3.0356825868480017e-05, | |
| "loss": 1.6384, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.3615423341669086, | |
| "learning_rate": 3.0212259068543837e-05, | |
| "loss": 1.4734, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.3863179074446679, | |
| "grad_norm": 0.3469703962175405, | |
| "learning_rate": 3.006750982215234e-05, | |
| "loss": 1.4386, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.391683433936955, | |
| "grad_norm": 0.3757992095280783, | |
| "learning_rate": 2.9922583196012037e-05, | |
| "loss": 1.5479, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.3970489604292422, | |
| "grad_norm": 0.3696529318558144, | |
| "learning_rate": 2.9777484263038306e-05, | |
| "loss": 1.4613, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.4024144869215291, | |
| "grad_norm": 0.36763443681114544, | |
| "learning_rate": 2.9632218102177862e-05, | |
| "loss": 1.4707, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.4077800134138163, | |
| "grad_norm": 0.3330227260573098, | |
| "learning_rate": 2.9486789798230917e-05, | |
| "loss": 1.5196, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.4131455399061033, | |
| "grad_norm": 0.3401075705229897, | |
| "learning_rate": 2.9341204441673266e-05, | |
| "loss": 1.5713, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.4185110663983904, | |
| "grad_norm": 0.38589937071982083, | |
| "learning_rate": 2.9195467128478044e-05, | |
| "loss": 1.5658, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.4238765928906774, | |
| "grad_norm": 0.36187499541792045, | |
| "learning_rate": 2.9049582959937392e-05, | |
| "loss": 1.5645, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.4292421193829645, | |
| "grad_norm": 0.3727884692802974, | |
| "learning_rate": 2.8903557042483887e-05, | |
| "loss": 1.5195, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.4346076458752515, | |
| "grad_norm": 0.3423144867083204, | |
| "learning_rate": 2.875739448751176e-05, | |
| "loss": 1.4897, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.4399731723675386, | |
| "grad_norm": 0.3809610047831763, | |
| "learning_rate": 2.8611100411198037e-05, | |
| "loss": 1.6107, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.4453386988598256, | |
| "grad_norm": 0.3466660038008262, | |
| "learning_rate": 2.8464679934323424e-05, | |
| "loss": 1.5203, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.4507042253521127, | |
| "grad_norm": 0.3858219936970091, | |
| "learning_rate": 2.8318138182093052e-05, | |
| "loss": 1.5211, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4560697518443997, | |
| "grad_norm": 0.3457606288538911, | |
| "learning_rate": 2.8171480283957118e-05, | |
| "loss": 1.4825, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.4614352783366868, | |
| "grad_norm": 0.4097169980235964, | |
| "learning_rate": 2.80247113734313e-05, | |
| "loss": 1.5276, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.4668008048289738, | |
| "grad_norm": 0.39937404860513825, | |
| "learning_rate": 2.7877836587917072e-05, | |
| "loss": 1.5022, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.472166331321261, | |
| "grad_norm": 0.36070057912441533, | |
| "learning_rate": 2.773086106852192e-05, | |
| "loss": 1.4587, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.477531857813548, | |
| "grad_norm": 0.4013616820391148, | |
| "learning_rate": 2.7583789959879303e-05, | |
| "loss": 1.5908, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.482897384305835, | |
| "grad_norm": 0.4150531718746046, | |
| "learning_rate": 2.7436628409968664e-05, | |
| "loss": 1.4511, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.488262910798122, | |
| "grad_norm": 0.38578622280087393, | |
| "learning_rate": 2.728938156993517e-05, | |
| "loss": 1.5407, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.4936284372904092, | |
| "grad_norm": 0.3844783258031986, | |
| "learning_rate": 2.7142054593909422e-05, | |
| "loss": 1.5349, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.4989939637826961, | |
| "grad_norm": 0.4144431176495637, | |
| "learning_rate": 2.6994652638827078e-05, | |
| "loss": 1.602, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.5043594902749833, | |
| "grad_norm": 0.402711914294424, | |
| "learning_rate": 2.6847180864248283e-05, | |
| "loss": 1.5902, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.5097250167672702, | |
| "grad_norm": 0.3378799881196754, | |
| "learning_rate": 2.6699644432177112e-05, | |
| "loss": 1.5514, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.5150905432595574, | |
| "grad_norm": 0.4113537256498264, | |
| "learning_rate": 2.655204850688085e-05, | |
| "loss": 1.4614, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.5150905432595574, | |
| "eval_loss": 2.1373159885406494, | |
| "eval_runtime": 403.7482, | |
| "eval_samples_per_second": 6.313, | |
| "eval_steps_per_second": 0.79, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.5204560697518446, | |
| "grad_norm": 0.3645339916373801, | |
| "learning_rate": 2.6404398254709284e-05, | |
| "loss": 1.4716, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.5258215962441315, | |
| "grad_norm": 0.3662916121671904, | |
| "learning_rate": 2.625669884391377e-05, | |
| "loss": 1.5145, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.5311871227364184, | |
| "grad_norm": 0.4062086365442157, | |
| "learning_rate": 2.610895544446641e-05, | |
| "loss": 1.5513, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.5365526492287056, | |
| "grad_norm": 0.3693755738805308, | |
| "learning_rate": 2.596117322787907e-05, | |
| "loss": 1.5498, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.5419181757209928, | |
| "grad_norm": 0.3589026792727439, | |
| "learning_rate": 2.5813357367022305e-05, | |
| "loss": 1.5211, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.5472837022132797, | |
| "grad_norm": 0.38017444587432575, | |
| "learning_rate": 2.566551303594437e-05, | |
| "loss": 1.5342, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.5526492287055667, | |
| "grad_norm": 0.3537300073106448, | |
| "learning_rate": 2.551764540969005e-05, | |
| "loss": 1.5109, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.5580147551978538, | |
| "grad_norm": 0.3752836981822812, | |
| "learning_rate": 2.5369759664119537e-05, | |
| "loss": 1.5015, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.563380281690141, | |
| "grad_norm": 0.3810876452684417, | |
| "learning_rate": 2.5221860975727275e-05, | |
| "loss": 1.5686, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.568745808182428, | |
| "grad_norm": 0.3688966465073692, | |
| "learning_rate": 2.5073954521460745e-05, | |
| "loss": 1.5666, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.5741113346747149, | |
| "grad_norm": 0.3740373962224673, | |
| "learning_rate": 2.4926045478539257e-05, | |
| "loss": 1.5541, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.579476861167002, | |
| "grad_norm": 0.4415523054571514, | |
| "learning_rate": 2.4778139024272724e-05, | |
| "loss": 1.62, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.5848423876592892, | |
| "grad_norm": 0.3291219023047839, | |
| "learning_rate": 2.4630240335880462e-05, | |
| "loss": 1.5, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.5902079141515761, | |
| "grad_norm": 0.39622699819562734, | |
| "learning_rate": 2.4482354590309962e-05, | |
| "loss": 1.5358, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.595573440643863, | |
| "grad_norm": 0.35147156988608064, | |
| "learning_rate": 2.433448696405563e-05, | |
| "loss": 1.469, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.6009389671361502, | |
| "grad_norm": 0.3458967842322882, | |
| "learning_rate": 2.4186642632977697e-05, | |
| "loss": 1.5368, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.6063044936284374, | |
| "grad_norm": 0.45919555248078225, | |
| "learning_rate": 2.4038826772120932e-05, | |
| "loss": 1.6126, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.6116700201207244, | |
| "grad_norm": 0.397298560652149, | |
| "learning_rate": 2.3891044555533588e-05, | |
| "loss": 1.5273, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6170355466130113, | |
| "grad_norm": 0.33950881912700964, | |
| "learning_rate": 2.3743301156086244e-05, | |
| "loss": 1.5844, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.6224010731052985, | |
| "grad_norm": 0.3685147478174407, | |
| "learning_rate": 2.359560174529073e-05, | |
| "loss": 1.5306, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.6277665995975856, | |
| "grad_norm": 0.3668847123777601, | |
| "learning_rate": 2.3447951493119152e-05, | |
| "loss": 1.5114, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.6331321260898726, | |
| "grad_norm": 0.3532238202377609, | |
| "learning_rate": 2.3300355567822897e-05, | |
| "loss": 1.5568, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.6384976525821595, | |
| "grad_norm": 0.3810645993344603, | |
| "learning_rate": 2.3152819135751722e-05, | |
| "loss": 1.5121, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.6438631790744467, | |
| "grad_norm": 0.37909917013988087, | |
| "learning_rate": 2.300534736117292e-05, | |
| "loss": 1.4904, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.6492287055667338, | |
| "grad_norm": 0.38885366059907706, | |
| "learning_rate": 2.285794540609058e-05, | |
| "loss": 1.5852, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.6545942320590208, | |
| "grad_norm": 0.3657756765079751, | |
| "learning_rate": 2.2710618430064843e-05, | |
| "loss": 1.5529, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.6599597585513077, | |
| "grad_norm": 0.3397163303578702, | |
| "learning_rate": 2.256337159003134e-05, | |
| "loss": 1.4903, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.665325285043595, | |
| "grad_norm": 0.39777082501030225, | |
| "learning_rate": 2.2416210040120703e-05, | |
| "loss": 1.5159, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.670690811535882, | |
| "grad_norm": 0.3658755281773237, | |
| "learning_rate": 2.2269138931478084e-05, | |
| "loss": 1.5848, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.676056338028169, | |
| "grad_norm": 0.42665798217406753, | |
| "learning_rate": 2.2122163412082927e-05, | |
| "loss": 1.6133, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.681421864520456, | |
| "grad_norm": 0.40969972656814363, | |
| "learning_rate": 2.1975288626568713e-05, | |
| "loss": 1.6264, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.686787391012743, | |
| "grad_norm": 0.3283642172163527, | |
| "learning_rate": 2.1828519716042888e-05, | |
| "loss": 1.4812, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.6921529175050303, | |
| "grad_norm": 0.4058354316614303, | |
| "learning_rate": 2.1681861817906954e-05, | |
| "loss": 1.4827, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.6975184439973172, | |
| "grad_norm": 0.3466585039953347, | |
| "learning_rate": 2.153532006567658e-05, | |
| "loss": 1.5768, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.7028839704896042, | |
| "grad_norm": 0.33329283442783797, | |
| "learning_rate": 2.1388899588801965e-05, | |
| "loss": 1.577, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.7082494969818913, | |
| "grad_norm": 0.3513005805125947, | |
| "learning_rate": 2.1242605512488248e-05, | |
| "loss": 1.5835, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.7136150234741785, | |
| "grad_norm": 0.3294934417641562, | |
| "learning_rate": 2.109644295751612e-05, | |
| "loss": 1.4758, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.7189805499664654, | |
| "grad_norm": 0.33415233723527016, | |
| "learning_rate": 2.095041704006261e-05, | |
| "loss": 1.4853, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.7243460764587524, | |
| "grad_norm": 0.3283653112019817, | |
| "learning_rate": 2.080453287152196e-05, | |
| "loss": 1.4554, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.7297116029510395, | |
| "grad_norm": 0.32083254638693504, | |
| "learning_rate": 2.0658795558326743e-05, | |
| "loss": 1.5284, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.7350771294433267, | |
| "grad_norm": 0.37794249417672565, | |
| "learning_rate": 2.0513210201769085e-05, | |
| "loss": 1.5521, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.7404426559356136, | |
| "grad_norm": 0.3431657733658665, | |
| "learning_rate": 2.0367781897822147e-05, | |
| "loss": 1.5067, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.7458081824279006, | |
| "grad_norm": 0.33934139214309345, | |
| "learning_rate": 2.0222515736961696e-05, | |
| "loss": 1.4801, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.7511737089201878, | |
| "grad_norm": 0.3795356143714155, | |
| "learning_rate": 2.0077416803987965e-05, | |
| "loss": 1.6264, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.756539235412475, | |
| "grad_norm": 0.3507795265759812, | |
| "learning_rate": 1.993249017784766e-05, | |
| "loss": 1.4967, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.3648489232026123, | |
| "learning_rate": 1.9787740931456165e-05, | |
| "loss": 1.5561, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.7672702883970488, | |
| "grad_norm": 0.39626816117060937, | |
| "learning_rate": 1.9643174131519986e-05, | |
| "loss": 1.6362, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.7672702883970488, | |
| "eval_loss": 2.1242423057556152, | |
| "eval_runtime": 403.1029, | |
| "eval_samples_per_second": 6.323, | |
| "eval_steps_per_second": 0.791, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.772635814889336, | |
| "grad_norm": 0.35313890178416246, | |
| "learning_rate": 1.949879483835939e-05, | |
| "loss": 1.5147, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.7780013413816231, | |
| "grad_norm": 0.3269413977690858, | |
| "learning_rate": 1.935460810573127e-05, | |
| "loss": 1.5032, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.78336686787391, | |
| "grad_norm": 0.3521215800086969, | |
| "learning_rate": 1.9210618980652277e-05, | |
| "loss": 1.5734, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.788732394366197, | |
| "grad_norm": 0.3813859592452455, | |
| "learning_rate": 1.9066832503222128e-05, | |
| "loss": 1.4488, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.7940979208584842, | |
| "grad_norm": 0.31690523060028536, | |
| "learning_rate": 1.892325370644721e-05, | |
| "loss": 1.432, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.7994634473507714, | |
| "grad_norm": 0.3364023400710833, | |
| "learning_rate": 1.8779887616064383e-05, | |
| "loss": 1.4871, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.8048289738430583, | |
| "grad_norm": 0.3350534986332929, | |
| "learning_rate": 1.863673925036506e-05, | |
| "loss": 1.5055, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.8101945003353455, | |
| "grad_norm": 0.33398900975140067, | |
| "learning_rate": 1.8493813620019594e-05, | |
| "loss": 1.5005, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.8155600268276326, | |
| "grad_norm": 0.31906353499540524, | |
| "learning_rate": 1.835111572790183e-05, | |
| "loss": 1.4626, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.8209255533199196, | |
| "grad_norm": 0.34716461667960885, | |
| "learning_rate": 1.8208650568914033e-05, | |
| "loss": 1.509, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.8262910798122065, | |
| "grad_norm": 0.34582309441917697, | |
| "learning_rate": 1.8066423129812027e-05, | |
| "loss": 1.4824, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.8316566063044937, | |
| "grad_norm": 0.32325626999744855, | |
| "learning_rate": 1.792443838903065e-05, | |
| "loss": 1.4303, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.8370221327967808, | |
| "grad_norm": 0.32008576746121775, | |
| "learning_rate": 1.778270131650948e-05, | |
| "loss": 1.6245, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.8423876592890678, | |
| "grad_norm": 0.37457992451894306, | |
| "learning_rate": 1.7641216873518878e-05, | |
| "loss": 1.4618, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.8477531857813547, | |
| "grad_norm": 0.35750145316045384, | |
| "learning_rate": 1.749999001248635e-05, | |
| "loss": 1.5959, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.8531187122736419, | |
| "grad_norm": 0.3306845841796733, | |
| "learning_rate": 1.735902567682315e-05, | |
| "loss": 1.5816, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.858484238765929, | |
| "grad_norm": 0.38970827119176255, | |
| "learning_rate": 1.7218328800751288e-05, | |
| "loss": 1.5529, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.863849765258216, | |
| "grad_norm": 0.3217477415509899, | |
| "learning_rate": 1.7077904309130782e-05, | |
| "loss": 1.5559, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.869215291750503, | |
| "grad_norm": 0.36288548537121584, | |
| "learning_rate": 1.6937757117287278e-05, | |
| "loss": 1.4984, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.87458081824279, | |
| "grad_norm": 0.3460179189739247, | |
| "learning_rate": 1.6797892130840036e-05, | |
| "loss": 1.5385, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.8799463447350773, | |
| "grad_norm": 0.31554481787674976, | |
| "learning_rate": 1.665831424553015e-05, | |
| "loss": 1.504, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.8853118712273642, | |
| "grad_norm": 0.30549498763968225, | |
| "learning_rate": 1.651902834704924e-05, | |
| "loss": 1.5228, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.8906773977196512, | |
| "grad_norm": 0.3277064150411407, | |
| "learning_rate": 1.6380039310868416e-05, | |
| "loss": 1.4852, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.8960429242119383, | |
| "grad_norm": 0.31968263622304366, | |
| "learning_rate": 1.624135200206759e-05, | |
| "loss": 1.4611, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.9014084507042255, | |
| "grad_norm": 0.32902733969646153, | |
| "learning_rate": 1.6102971275165228e-05, | |
| "loss": 1.4833, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.9067739771965124, | |
| "grad_norm": 0.3299428742675132, | |
| "learning_rate": 1.596490197394841e-05, | |
| "loss": 1.4439, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.9121395036887994, | |
| "grad_norm": 0.34500427705894593, | |
| "learning_rate": 1.5827148931303277e-05, | |
| "loss": 1.5616, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.9175050301810865, | |
| "grad_norm": 0.32458276939765035, | |
| "learning_rate": 1.5689716969045848e-05, | |
| "loss": 1.4334, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.9228705566733737, | |
| "grad_norm": 0.3342429982300284, | |
| "learning_rate": 1.5552610897753292e-05, | |
| "loss": 1.5311, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.9282360831656606, | |
| "grad_norm": 0.3483438918440925, | |
| "learning_rate": 1.5415835516595465e-05, | |
| "loss": 1.4642, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.9336016096579476, | |
| "grad_norm": 0.324692472057597, | |
| "learning_rate": 1.5279395613166986e-05, | |
| "loss": 1.5336, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.9389671361502347, | |
| "grad_norm": 0.4011901483013197, | |
| "learning_rate": 1.5143295963319643e-05, | |
| "loss": 1.5634, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.944332662642522, | |
| "grad_norm": 0.3791986336861211, | |
| "learning_rate": 1.5007541330995197e-05, | |
| "loss": 1.5722, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.9496981891348089, | |
| "grad_norm": 0.3318324328550346, | |
| "learning_rate": 1.4872136468058661e-05, | |
| "loss": 1.6214, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.9550637156270958, | |
| "grad_norm": 0.3865281831290857, | |
| "learning_rate": 1.4737086114131943e-05, | |
| "loss": 1.548, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.960429242119383, | |
| "grad_norm": 0.3317269613634914, | |
| "learning_rate": 1.4602394996427942e-05, | |
| "loss": 1.5024, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.9657947686116701, | |
| "grad_norm": 0.36387655313468187, | |
| "learning_rate": 1.4468067829585108e-05, | |
| "loss": 1.5256, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.971160295103957, | |
| "grad_norm": 0.3494371010379711, | |
| "learning_rate": 1.4334109315502395e-05, | |
| "loss": 1.5559, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.976525821596244, | |
| "grad_norm": 0.36135083308665883, | |
| "learning_rate": 1.4200524143174677e-05, | |
| "loss": 1.5669, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.9818913480885312, | |
| "grad_norm": 0.34824673781683024, | |
| "learning_rate": 1.4067316988528617e-05, | |
| "loss": 1.6642, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.9872568745808183, | |
| "grad_norm": 0.36134664903507996, | |
| "learning_rate": 1.3934492514259003e-05, | |
| "loss": 1.5003, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.9926224010731053, | |
| "grad_norm": 0.3315306163140975, | |
| "learning_rate": 1.3802055369665534e-05, | |
| "loss": 1.3959, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.9979879275653922, | |
| "grad_norm": 0.3179771784326003, | |
| "learning_rate": 1.3670010190490073e-05, | |
| "loss": 1.5285, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 2.005365526492287, | |
| "grad_norm": 0.9915595865019317, | |
| "learning_rate": 1.3538361598754384e-05, | |
| "loss": 2.456, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 2.0107310529845743, | |
| "grad_norm": 0.6291566214590382, | |
| "learning_rate": 1.3407114202598369e-05, | |
| "loss": 1.1708, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.0160965794768613, | |
| "grad_norm": 0.5593895609190137, | |
| "learning_rate": 1.327627259611873e-05, | |
| "loss": 1.1864, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.021462105969148, | |
| "grad_norm": 0.4205563210949494, | |
| "learning_rate": 1.314584135920815e-05, | |
| "loss": 1.1374, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.021462105969148, | |
| "eval_loss": 2.232717752456665, | |
| "eval_runtime": 403.1787, | |
| "eval_samples_per_second": 6.322, | |
| "eval_steps_per_second": 0.791, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 2.026827632461435, | |
| "grad_norm": 0.7137961053387198, | |
| "learning_rate": 1.301582505739506e-05, | |
| "loss": 1.2243, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 2.0321931589537225, | |
| "grad_norm": 0.7110032999802158, | |
| "learning_rate": 1.2886228241683749e-05, | |
| "loss": 1.1914, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 2.0375586854460095, | |
| "grad_norm": 0.6070986317073753, | |
| "learning_rate": 1.2757055448395092e-05, | |
| "loss": 1.134, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 2.0429242119382964, | |
| "grad_norm": 0.4599660060017544, | |
| "learning_rate": 1.2628311199007764e-05, | |
| "loss": 1.1466, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.0482897384305834, | |
| "grad_norm": 0.48013581218581647, | |
| "learning_rate": 1.2500000000000006e-05, | |
| "loss": 1.2032, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 2.0536552649228708, | |
| "grad_norm": 0.47190986939626134, | |
| "learning_rate": 1.2372126342691798e-05, | |
| "loss": 1.1103, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 2.0590207914151577, | |
| "grad_norm": 0.4474749587522701, | |
| "learning_rate": 1.2244694703087728e-05, | |
| "loss": 1.1502, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 2.0643863179074446, | |
| "grad_norm": 0.44999362516877806, | |
| "learning_rate": 1.2117709541720306e-05, | |
| "loss": 1.1528, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 2.0697518443997316, | |
| "grad_norm": 0.45904376696237503, | |
| "learning_rate": 1.1991175303493793e-05, | |
| "loss": 1.2056, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 2.075117370892019, | |
| "grad_norm": 0.4344382184186286, | |
| "learning_rate": 1.1865096417528635e-05, | |
| "loss": 1.1107, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 2.080482897384306, | |
| "grad_norm": 0.3972755677260206, | |
| "learning_rate": 1.173947729700644e-05, | |
| "loss": 1.1499, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 2.085848423876593, | |
| "grad_norm": 0.3950446642940489, | |
| "learning_rate": 1.1614322339015484e-05, | |
| "loss": 1.1557, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 2.09121395036888, | |
| "grad_norm": 0.4447326020481244, | |
| "learning_rate": 1.1489635924396817e-05, | |
| "loss": 1.1555, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 2.096579476861167, | |
| "grad_norm": 0.419897228759811, | |
| "learning_rate": 1.1365422417590878e-05, | |
| "loss": 1.1756, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.101945003353454, | |
| "grad_norm": 0.41249904086804146, | |
| "learning_rate": 1.1241686166484805e-05, | |
| "loss": 1.1706, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 2.107310529845741, | |
| "grad_norm": 0.40775526734089124, | |
| "learning_rate": 1.1118431502260163e-05, | |
| "loss": 1.2389, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 2.112676056338028, | |
| "grad_norm": 0.41120972999065375, | |
| "learning_rate": 1.0995662739241347e-05, | |
| "loss": 1.2027, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 2.1180415828303154, | |
| "grad_norm": 0.36915354132829215, | |
| "learning_rate": 1.0873384174744641e-05, | |
| "loss": 1.1211, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 2.1234071093226023, | |
| "grad_norm": 0.3734288072829599, | |
| "learning_rate": 1.0751600088927713e-05, | |
| "loss": 1.2408, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.1287726358148893, | |
| "grad_norm": 0.40175568463326916, | |
| "learning_rate": 1.063031474463983e-05, | |
| "loss": 1.1744, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 2.1341381623071762, | |
| "grad_norm": 0.3674474877057401, | |
| "learning_rate": 1.0509532387272641e-05, | |
| "loss": 1.1628, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 2.1395036887994636, | |
| "grad_norm": 0.36762315854199107, | |
| "learning_rate": 1.0389257244611602e-05, | |
| "loss": 1.162, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 2.1448692152917506, | |
| "grad_norm": 0.3414954721940739, | |
| "learning_rate": 1.0269493526687915e-05, | |
| "loss": 1.182, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 2.1502347417840375, | |
| "grad_norm": 0.38061259662736135, | |
| "learning_rate": 1.0150245425631235e-05, | |
| "loss": 1.1561, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.1556002682763244, | |
| "grad_norm": 0.36884370471677075, | |
| "learning_rate": 1.0031517115522926e-05, | |
| "loss": 1.2107, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 2.160965794768612, | |
| "grad_norm": 0.3461756268853831, | |
| "learning_rate": 9.913312752249903e-06, | |
| "loss": 1.1919, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 2.166331321260899, | |
| "grad_norm": 0.3453066453102883, | |
| "learning_rate": 9.795636473359208e-06, | |
| "loss": 1.1735, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 2.1716968477531857, | |
| "grad_norm": 0.3585272022980633, | |
| "learning_rate": 9.678492397913167e-06, | |
| "loss": 1.2119, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 2.1770623742454727, | |
| "grad_norm": 0.34484629378614046, | |
| "learning_rate": 9.561884626345205e-06, | |
| "loss": 1.1628, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.18242790073776, | |
| "grad_norm": 0.340448631343189, | |
| "learning_rate": 9.445817240316332e-06, | |
| "loss": 1.1534, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 2.187793427230047, | |
| "grad_norm": 0.3491564049843763, | |
| "learning_rate": 9.330294302572242e-06, | |
| "loss": 1.1707, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 2.193158953722334, | |
| "grad_norm": 0.3374441017987585, | |
| "learning_rate": 9.215319856801158e-06, | |
| "loss": 1.124, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 2.198524480214621, | |
| "grad_norm": 0.33387420272144586, | |
| "learning_rate": 9.10089792749223e-06, | |
| "loss": 1.1547, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 2.2038900067069083, | |
| "grad_norm": 0.34377542398740535, | |
| "learning_rate": 8.987032519794666e-06, | |
| "loss": 1.1865, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.209255533199195, | |
| "grad_norm": 0.34410948368421596, | |
| "learning_rate": 8.873727619377611e-06, | |
| "loss": 1.1807, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 2.214621059691482, | |
| "grad_norm": 0.3454829815928742, | |
| "learning_rate": 8.760987192290557e-06, | |
| "loss": 1.2488, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 2.219986586183769, | |
| "grad_norm": 0.34035820590189597, | |
| "learning_rate": 8.648815184824544e-06, | |
| "loss": 1.1629, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 2.2253521126760565, | |
| "grad_norm": 0.35672991922416875, | |
| "learning_rate": 8.537215523374038e-06, | |
| "loss": 1.2238, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 2.2307176391683434, | |
| "grad_norm": 0.351148929709105, | |
| "learning_rate": 8.426192114299484e-06, | |
| "loss": 1.2074, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.2360831656606304, | |
| "grad_norm": 0.3449030405647957, | |
| "learning_rate": 8.315748843790563e-06, | |
| "loss": 1.188, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 2.2414486921529173, | |
| "grad_norm": 0.3432687950624104, | |
| "learning_rate": 8.20588957773018e-06, | |
| "loss": 1.1524, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 2.2468142186452047, | |
| "grad_norm": 0.34423459328901956, | |
| "learning_rate": 8.096618161559142e-06, | |
| "loss": 1.143, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 2.2521797451374916, | |
| "grad_norm": 0.32619680555917097, | |
| "learning_rate": 7.987938420141536e-06, | |
| "loss": 1.1641, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 2.2575452716297786, | |
| "grad_norm": 0.33552529458130786, | |
| "learning_rate": 7.879854157630861e-06, | |
| "loss": 1.1301, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.262910798122066, | |
| "grad_norm": 0.3336572936896806, | |
| "learning_rate": 7.772369157336874e-06, | |
| "loss": 1.119, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 2.268276324614353, | |
| "grad_norm": 0.33482717731077677, | |
| "learning_rate": 7.665487181593145e-06, | |
| "loss": 1.1412, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 2.27364185110664, | |
| "grad_norm": 0.31988515510077037, | |
| "learning_rate": 7.5592119716253855e-06, | |
| "loss": 1.2301, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.27364185110664, | |
| "eval_loss": 2.2243738174438477, | |
| "eval_runtime": 403.6162, | |
| "eval_samples_per_second": 6.315, | |
| "eval_steps_per_second": 0.79, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 2.279007377598927, | |
| "grad_norm": 0.32212835750220664, | |
| "learning_rate": 7.4535472474204645e-06, | |
| "loss": 1.1127, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 2.2843729040912137, | |
| "grad_norm": 0.3541133973544153, | |
| "learning_rate": 7.348496707596242e-06, | |
| "loss": 1.0762, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.289738430583501, | |
| "grad_norm": 0.3272644781347247, | |
| "learning_rate": 7.244064029272049e-06, | |
| "loss": 1.1938, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 2.295103957075788, | |
| "grad_norm": 0.3248701960652722, | |
| "learning_rate": 7.140252867939995e-06, | |
| "loss": 1.1952, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 2.300469483568075, | |
| "grad_norm": 0.3278958191476634, | |
| "learning_rate": 7.037066857337058e-06, | |
| "loss": 1.2259, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.3058350100603624, | |
| "grad_norm": 0.33588934698731865, | |
| "learning_rate": 6.934509609317821e-06, | |
| "loss": 1.0948, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 2.3112005365526493, | |
| "grad_norm": 0.33096764574513354, | |
| "learning_rate": 6.832584713728101e-06, | |
| "loss": 1.1655, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.3165660630449363, | |
| "grad_norm": 0.3363375955193017, | |
| "learning_rate": 6.7312957382792556e-06, | |
| "loss": 1.1786, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 2.3219315895372232, | |
| "grad_norm": 0.3384101006551099, | |
| "learning_rate": 6.6306462284233234e-06, | |
| "loss": 1.1761, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 2.32729711602951, | |
| "grad_norm": 0.6194898145769814, | |
| "learning_rate": 6.5306397072289114e-06, | |
| "loss": 1.1916, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 2.3326626425217976, | |
| "grad_norm": 0.33449734643167867, | |
| "learning_rate": 6.431279675257873e-06, | |
| "loss": 1.0894, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 2.3380281690140845, | |
| "grad_norm": 0.3449479370773172, | |
| "learning_rate": 6.332569610442807e-06, | |
| "loss": 1.2021, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.3433936955063714, | |
| "grad_norm": 0.3324865938198681, | |
| "learning_rate": 6.234512967965261e-06, | |
| "loss": 1.1583, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 2.348759221998659, | |
| "grad_norm": 0.3233624144923054, | |
| "learning_rate": 6.1371131801348434e-06, | |
| "loss": 1.1243, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 2.3541247484909458, | |
| "grad_norm": 0.34771192845713556, | |
| "learning_rate": 6.040373656269041e-06, | |
| "loss": 1.1667, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 2.3594902749832327, | |
| "grad_norm": 0.3476437588037759, | |
| "learning_rate": 5.944297782573918e-06, | |
| "loss": 1.1192, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 2.3648558014755197, | |
| "grad_norm": 0.36436230780329204, | |
| "learning_rate": 5.848888922025553e-06, | |
| "loss": 1.2521, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.3702213279678066, | |
| "grad_norm": 0.35739290718549466, | |
| "learning_rate": 5.75415041425234e-06, | |
| "loss": 1.1772, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 2.375586854460094, | |
| "grad_norm": 0.3187619530515648, | |
| "learning_rate": 5.660085575418114e-06, | |
| "loss": 1.0855, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 2.380952380952381, | |
| "grad_norm": 0.3402178296399225, | |
| "learning_rate": 5.566697698106024e-06, | |
| "loss": 1.1531, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 2.386317907444668, | |
| "grad_norm": 0.3428456906688321, | |
| "learning_rate": 5.473990051203298e-06, | |
| "loss": 1.1289, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 2.3916834339369553, | |
| "grad_norm": 0.3251361090262083, | |
| "learning_rate": 5.381965879786868e-06, | |
| "loss": 1.1864, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.397048960429242, | |
| "grad_norm": 0.3436986211314983, | |
| "learning_rate": 5.290628405009717e-06, | |
| "loss": 1.0949, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 2.402414486921529, | |
| "grad_norm": 0.3350393677646101, | |
| "learning_rate": 5.199980823988157e-06, | |
| "loss": 1.0944, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 2.407780013413816, | |
| "grad_norm": 0.32046838027326285, | |
| "learning_rate": 5.110026309689922e-06, | |
| "loss": 1.1519, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 2.4131455399061035, | |
| "grad_norm": 0.33210748691525244, | |
| "learning_rate": 5.020768010823102e-06, | |
| "loss": 1.216, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 2.4185110663983904, | |
| "grad_norm": 0.31198379909853313, | |
| "learning_rate": 4.932209051725914e-06, | |
| "loss": 1.0948, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.4238765928906774, | |
| "grad_norm": 0.3143147465067978, | |
| "learning_rate": 4.844352532257351e-06, | |
| "loss": 1.0641, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 2.4292421193829643, | |
| "grad_norm": 0.3299169974610766, | |
| "learning_rate": 4.757201527688693e-06, | |
| "loss": 1.1331, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 2.4346076458752517, | |
| "grad_norm": 0.3240731706047221, | |
| "learning_rate": 4.67075908859583e-06, | |
| "loss": 1.1424, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 2.4399731723675386, | |
| "grad_norm": 0.31726341227978605, | |
| "learning_rate": 4.585028240752498e-06, | |
| "loss": 1.0585, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 2.4453386988598256, | |
| "grad_norm": 0.3133109530668219, | |
| "learning_rate": 4.500011985024363e-06, | |
| "loss": 1.1124, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.4507042253521125, | |
| "grad_norm": 0.32063200368226685, | |
| "learning_rate": 4.415713297263987e-06, | |
| "loss": 1.127, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 2.4560697518444, | |
| "grad_norm": 0.3230470474940416, | |
| "learning_rate": 4.332135128206666e-06, | |
| "loss": 1.164, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 2.461435278336687, | |
| "grad_norm": 0.32502523603989114, | |
| "learning_rate": 4.2492804033671145e-06, | |
| "loss": 1.1437, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 2.466800804828974, | |
| "grad_norm": 0.3246398479701492, | |
| "learning_rate": 4.167152022937124e-06, | |
| "loss": 1.1795, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.4721663313212607, | |
| "grad_norm": 0.33392434019026146, | |
| "learning_rate": 4.085752861683981e-06, | |
| "loss": 1.1344, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.477531857813548, | |
| "grad_norm": 0.32290812833166643, | |
| "learning_rate": 4.005085768849856e-06, | |
| "loss": 1.1691, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.482897384305835, | |
| "grad_norm": 0.3351539620126592, | |
| "learning_rate": 3.925153568052123e-06, | |
| "loss": 1.1926, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.488262910798122, | |
| "grad_norm": 0.32575534244471027, | |
| "learning_rate": 3.845959057184453e-06, | |
| "loss": 1.185, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.493628437290409, | |
| "grad_norm": 0.3318857239335513, | |
| "learning_rate": 3.767505008318914e-06, | |
| "loss": 1.1056, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.4989939637826963, | |
| "grad_norm": 0.3332508682060839, | |
| "learning_rate": 3.689794167608937e-06, | |
| "loss": 1.0975, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.5043594902749833, | |
| "grad_norm": 0.3322015968073437, | |
| "learning_rate": 3.6128292551931924e-06, | |
| "loss": 1.1779, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.5097250167672702, | |
| "grad_norm": 0.3339421172566103, | |
| "learning_rate": 3.536612965100361e-06, | |
| "loss": 1.1456, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.515090543259557, | |
| "grad_norm": 0.32193009926200084, | |
| "learning_rate": 3.4611479651548457e-06, | |
| "loss": 1.1247, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.5204560697518446, | |
| "grad_norm": 0.3275332021298932, | |
| "learning_rate": 3.3864368968834074e-06, | |
| "loss": 1.1583, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.5258215962441315, | |
| "grad_norm": 0.3322580175015273, | |
| "learning_rate": 3.3124823754226624e-06, | |
| "loss": 1.1773, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.5258215962441315, | |
| "eval_loss": 2.2257003784179688, | |
| "eval_runtime": 403.3972, | |
| "eval_samples_per_second": 6.319, | |
| "eval_steps_per_second": 0.791, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.5311871227364184, | |
| "grad_norm": 0.3230969965574369, | |
| "learning_rate": 3.239286989427573e-06, | |
| "loss": 1.1593, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 2.536552649228706, | |
| "grad_norm": 0.3206756438379543, | |
| "learning_rate": 3.166853300980821e-06, | |
| "loss": 1.1583, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.5419181757209928, | |
| "grad_norm": 0.32211926813307257, | |
| "learning_rate": 3.095183845503144e-06, | |
| "loss": 1.1679, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 2.5472837022132797, | |
| "grad_norm": 0.326559743052197, | |
| "learning_rate": 3.024281131664569e-06, | |
| "loss": 1.1188, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.5526492287055667, | |
| "grad_norm": 0.3165584436088936, | |
| "learning_rate": 2.9541476412966036e-06, | |
| "loss": 1.1372, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.5580147551978536, | |
| "grad_norm": 0.31339823548327456, | |
| "learning_rate": 2.884785829305381e-06, | |
| "loss": 1.1235, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.563380281690141, | |
| "grad_norm": 0.33296469052198463, | |
| "learning_rate": 2.8161981235857143e-06, | |
| "loss": 1.0565, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 2.568745808182428, | |
| "grad_norm": 0.3222335104308849, | |
| "learning_rate": 2.7483869249360915e-06, | |
| "loss": 1.1068, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.574111334674715, | |
| "grad_norm": 0.3123105201965395, | |
| "learning_rate": 2.681354606974698e-06, | |
| "loss": 1.1675, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 2.5794768611670023, | |
| "grad_norm": 0.33382708976050407, | |
| "learning_rate": 2.615103516056275e-06, | |
| "loss": 1.1605, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.584842387659289, | |
| "grad_norm": 0.3218837631971954, | |
| "learning_rate": 2.5496359711900118e-06, | |
| "loss": 1.1689, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 2.590207914151576, | |
| "grad_norm": 0.31291467673725704, | |
| "learning_rate": 2.484954263958383e-06, | |
| "loss": 1.1812, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.595573440643863, | |
| "grad_norm": 0.32492352260184354, | |
| "learning_rate": 2.4210606584369104e-06, | |
| "loss": 1.1555, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 2.60093896713615, | |
| "grad_norm": 0.3504213880385304, | |
| "learning_rate": 2.35795739111494e-06, | |
| "loss": 1.1061, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.6063044936284374, | |
| "grad_norm": 0.31100084169291436, | |
| "learning_rate": 2.2956466708173307e-06, | |
| "loss": 1.0928, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.6116700201207244, | |
| "grad_norm": 0.3248255726906159, | |
| "learning_rate": 2.2341306786271695e-06, | |
| "loss": 1.1183, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.6170355466130113, | |
| "grad_norm": 0.31866373634955536, | |
| "learning_rate": 2.173411567809394e-06, | |
| "loss": 1.1753, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 2.6224010731052987, | |
| "grad_norm": 0.3352823580996406, | |
| "learning_rate": 2.113491463735437e-06, | |
| "loss": 1.1207, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.6277665995975856, | |
| "grad_norm": 0.31359112976859854, | |
| "learning_rate": 2.0543724638088347e-06, | |
| "loss": 1.1801, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 2.6331321260898726, | |
| "grad_norm": 0.33926000212154844, | |
| "learning_rate": 1.996056637391805e-06, | |
| "loss": 1.2241, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.6384976525821595, | |
| "grad_norm": 0.3371162704533217, | |
| "learning_rate": 1.938546025732807e-06, | |
| "loss": 1.0681, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.6438631790744465, | |
| "grad_norm": 0.3190932843505867, | |
| "learning_rate": 1.881842641895104e-06, | |
| "loss": 1.1484, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.649228705566734, | |
| "grad_norm": 0.3275133328320819, | |
| "learning_rate": 1.8259484706862951e-06, | |
| "loss": 1.1495, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.654594232059021, | |
| "grad_norm": 0.32869987839977655, | |
| "learning_rate": 1.7708654685888337e-06, | |
| "loss": 1.175, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.6599597585513077, | |
| "grad_norm": 0.3185931737498036, | |
| "learning_rate": 1.7165955636915392e-06, | |
| "loss": 1.071, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.665325285043595, | |
| "grad_norm": 0.31601212994979094, | |
| "learning_rate": 1.6631406556221334e-06, | |
| "loss": 1.1135, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.670690811535882, | |
| "grad_norm": 0.3252786042303215, | |
| "learning_rate": 1.6105026154807218e-06, | |
| "loss": 1.0958, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.676056338028169, | |
| "grad_norm": 0.3613550247333106, | |
| "learning_rate": 1.558683285774304e-06, | |
| "loss": 1.1663, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.681421864520456, | |
| "grad_norm": 0.3343048936672064, | |
| "learning_rate": 1.5076844803522922e-06, | |
| "loss": 1.2131, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.686787391012743, | |
| "grad_norm": 0.31812219422372123, | |
| "learning_rate": 1.457507984343001e-06, | |
| "loss": 1.2014, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.6921529175050303, | |
| "grad_norm": 0.31499628329755797, | |
| "learning_rate": 1.408155554091184e-06, | |
| "loss": 1.1767, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.697518443997317, | |
| "grad_norm": 0.3218866465290174, | |
| "learning_rate": 1.3596289170965309e-06, | |
| "loss": 1.1169, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.702883970489604, | |
| "grad_norm": 0.32278919495138014, | |
| "learning_rate": 1.3119297719532242e-06, | |
| "loss": 1.1428, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.7082494969818915, | |
| "grad_norm": 0.3249191581938583, | |
| "learning_rate": 1.265059788290468e-06, | |
| "loss": 1.1663, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.7136150234741785, | |
| "grad_norm": 0.3252541633067707, | |
| "learning_rate": 1.2190206067140442e-06, | |
| "loss": 1.2813, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.7189805499664654, | |
| "grad_norm": 0.3106253606000537, | |
| "learning_rate": 1.173813838748894e-06, | |
| "loss": 1.1415, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.7243460764587524, | |
| "grad_norm": 0.314646801186305, | |
| "learning_rate": 1.129441066782702e-06, | |
| "loss": 1.1541, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.7297116029510393, | |
| "grad_norm": 0.3213537454716302, | |
| "learning_rate": 1.0859038440105164e-06, | |
| "loss": 1.138, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.7350771294433267, | |
| "grad_norm": 0.32136955676763673, | |
| "learning_rate": 1.0432036943803708e-06, | |
| "loss": 1.1231, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.7404426559356136, | |
| "grad_norm": 0.3243563096847065, | |
| "learning_rate": 1.0013421125399519e-06, | |
| "loss": 1.244, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.7458081824279006, | |
| "grad_norm": 0.3089576646257491, | |
| "learning_rate": 9.6032056378427e-07, | |
| "loss": 1.1554, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.751173708920188, | |
| "grad_norm": 0.3095090303092117, | |
| "learning_rate": 9.201404840043726e-07, | |
| "loss": 1.1231, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.756539235412475, | |
| "grad_norm": 0.3147170138443251, | |
| "learning_rate": 8.808032796371019e-07, | |
| "loss": 1.0902, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.761904761904762, | |
| "grad_norm": 0.29507123132031615, | |
| "learning_rate": 8.423103276158306e-07, | |
| "loss": 1.0369, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.767270288397049, | |
| "grad_norm": 0.30105732124479273, | |
| "learning_rate": 8.046629753222957e-07, | |
| "loss": 1.1041, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.7726358148893357, | |
| "grad_norm": 0.3224663865586402, | |
| "learning_rate": 7.678625405394157e-07, | |
| "loss": 1.1015, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.778001341381623, | |
| "grad_norm": 0.3222742299894699, | |
| "learning_rate": 7.319103114051707e-07, | |
| "loss": 1.1864, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.778001341381623, | |
| "eval_loss": 2.225644826889038, | |
| "eval_runtime": 403.2371, | |
| "eval_samples_per_second": 6.321, | |
| "eval_steps_per_second": 0.791, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.78336686787391, | |
| "grad_norm": 0.3171003364088583, | |
| "learning_rate": 6.9680754636752e-07, | |
| "loss": 1.1892, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.788732394366197, | |
| "grad_norm": 0.3119059918714888, | |
| "learning_rate": 6.625554741403333e-07, | |
| "loss": 1.2042, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.7940979208584844, | |
| "grad_norm": 0.3203009443769252, | |
| "learning_rate": 6.291552936604134e-07, | |
| "loss": 1.1335, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.7994634473507714, | |
| "grad_norm": 0.3288968824749739, | |
| "learning_rate": 5.966081740454932e-07, | |
| "loss": 1.192, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.8048289738430583, | |
| "grad_norm": 0.31460862008525864, | |
| "learning_rate": 5.649152545533332e-07, | |
| "loss": 1.1729, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.8101945003353457, | |
| "grad_norm": 0.3217212545480734, | |
| "learning_rate": 5.340776445418472e-07, | |
| "loss": 1.1193, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.8155600268276326, | |
| "grad_norm": 0.3123123100906072, | |
| "learning_rate": 5.040964234302559e-07, | |
| "loss": 1.133, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.8209255533199196, | |
| "grad_norm": 0.31788975020445803, | |
| "learning_rate": 4.749726406613142e-07, | |
| "loss": 1.1569, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.8262910798122065, | |
| "grad_norm": 0.35427580228034644, | |
| "learning_rate": 4.4670731566457126e-07, | |
| "loss": 1.185, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.8316566063044935, | |
| "grad_norm": 0.3078993071472304, | |
| "learning_rate": 4.1930143782070444e-07, | |
| "loss": 1.245, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.837022132796781, | |
| "grad_norm": 0.627826963441227, | |
| "learning_rate": 3.9275596642685543e-07, | |
| "loss": 1.1768, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.842387659289068, | |
| "grad_norm": 0.3180386319315376, | |
| "learning_rate": 3.670718306630766e-07, | |
| "loss": 1.1915, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.8477531857813547, | |
| "grad_norm": 0.3183014815220175, | |
| "learning_rate": 3.422499295598069e-07, | |
| "loss": 1.1618, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.853118712273642, | |
| "grad_norm": 0.31162900762652046, | |
| "learning_rate": 3.1829113196638614e-07, | |
| "loss": 1.1459, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.858484238765929, | |
| "grad_norm": 0.31454238955052416, | |
| "learning_rate": 2.9519627652065674e-07, | |
| "loss": 1.1396, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.863849765258216, | |
| "grad_norm": 0.33536803393906284, | |
| "learning_rate": 2.7296617161960415e-07, | |
| "loss": 1.1325, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.869215291750503, | |
| "grad_norm": 0.3118341382685939, | |
| "learning_rate": 2.5160159539105443e-07, | |
| "loss": 1.186, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.87458081824279, | |
| "grad_norm": 0.30550007371594207, | |
| "learning_rate": 2.3110329566645162e-07, | |
| "loss": 1.1615, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.8799463447350773, | |
| "grad_norm": 0.3290186323857588, | |
| "learning_rate": 2.1147198995466467e-07, | |
| "loss": 1.1199, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.885311871227364, | |
| "grad_norm": 0.30975536579912843, | |
| "learning_rate": 1.927083654168854e-07, | |
| "loss": 1.1207, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.890677397719651, | |
| "grad_norm": 0.3272675259115993, | |
| "learning_rate": 1.7481307884256727e-07, | |
| "loss": 1.1211, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.8960429242119385, | |
| "grad_norm": 0.3186123525466991, | |
| "learning_rate": 1.5778675662643793e-07, | |
| "loss": 1.2159, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.9014084507042255, | |
| "grad_norm": 0.3238665460415356, | |
| "learning_rate": 1.4162999474657268e-07, | |
| "loss": 1.1226, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.9067739771965124, | |
| "grad_norm": 0.300927276396165, | |
| "learning_rate": 1.2634335874353585e-07, | |
| "loss": 1.1584, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.9121395036887994, | |
| "grad_norm": 0.3225080409751762, | |
| "learning_rate": 1.1192738370058575e-07, | |
| "loss": 1.1549, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.9175050301810863, | |
| "grad_norm": 0.3014139103111567, | |
| "learning_rate": 9.838257422493668e-08, | |
| "loss": 1.0577, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.9228705566733737, | |
| "grad_norm": 0.33343711969938683, | |
| "learning_rate": 8.570940443010655e-08, | |
| "loss": 1.1204, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.9282360831656606, | |
| "grad_norm": 0.3001682609568673, | |
| "learning_rate": 7.390831791931896e-08, | |
| "loss": 1.116, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.9336016096579476, | |
| "grad_norm": 0.3132232444134447, | |
| "learning_rate": 6.297972776996286e-08, | |
| "loss": 1.1668, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.938967136150235, | |
| "grad_norm": 0.317135513260303, | |
| "learning_rate": 5.2924016519156906e-08, | |
| "loss": 1.1322, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.944332662642522, | |
| "grad_norm": 0.324986722470258, | |
| "learning_rate": 4.374153615033794e-08, | |
| "loss": 1.1091, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.949698189134809, | |
| "grad_norm": 0.30922413089636414, | |
| "learning_rate": 3.543260808095139e-08, | |
| "loss": 1.141, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.955063715627096, | |
| "grad_norm": 0.31855167384942973, | |
| "learning_rate": 2.799752315119919e-08, | |
| "loss": 1.2269, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.9604292421193827, | |
| "grad_norm": 0.3074014352539391, | |
| "learning_rate": 2.1436541613853444e-08, | |
| "loss": 1.1442, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.96579476861167, | |
| "grad_norm": 0.3064677206528894, | |
| "learning_rate": 1.5749893125160954e-08, | |
| "loss": 1.1533, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.971160295103957, | |
| "grad_norm": 0.3049690197063334, | |
| "learning_rate": 1.0937776736782978e-08, | |
| "loss": 1.112, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.976525821596244, | |
| "grad_norm": 0.3030005645927827, | |
| "learning_rate": 7.00036088885081e-09, | |
| "loss": 1.1847, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.9818913480885314, | |
| "grad_norm": 0.32049007711358624, | |
| "learning_rate": 3.9377834040538185e-09, | |
| "loss": 1.1018, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.9872568745808183, | |
| "grad_norm": 0.3209495594839988, | |
| "learning_rate": 1.7501514828183185e-09, | |
| "loss": 1.1483, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.9926224010731053, | |
| "grad_norm": 0.3215801377358293, | |
| "learning_rate": 4.3754169955778634e-10, | |
| "loss": 1.1563, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.9979879275653922, | |
| "grad_norm": 0.29931532757386853, | |
| "learning_rate": 0.0, | |
| "loss": 1.1306, | |
| "step": 558 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 558, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 47, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4979627422066934e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |