| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.49438202247191, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011235955056179775, | |
| "grad_norm": 3.1517910957336426, | |
| "learning_rate": 2.2222222222222224e-07, | |
| "loss": 0.4045, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02247191011235955, | |
| "grad_norm": 3.8104264736175537, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 0.4326, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.033707865168539325, | |
| "grad_norm": 3.410100221633911, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.4196, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0449438202247191, | |
| "grad_norm": 3.2662692070007324, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 0.3744, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.056179775280898875, | |
| "grad_norm": 3.052116632461548, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 0.4275, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06741573033707865, | |
| "grad_norm": 2.697885513305664, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.3805, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07865168539325842, | |
| "grad_norm": 2.9053075313568115, | |
| "learning_rate": 1.5555555555555558e-06, | |
| "loss": 0.3949, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0898876404494382, | |
| "grad_norm": 2.0928916931152344, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 0.3856, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.10112359550561797, | |
| "grad_norm": 1.903058409690857, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.3749, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.11235955056179775, | |
| "grad_norm": 1.6796547174453735, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.3811, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12359550561797752, | |
| "grad_norm": 1.761483907699585, | |
| "learning_rate": 2.4444444444444447e-06, | |
| "loss": 0.3902, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1348314606741573, | |
| "grad_norm": 1.6522550582885742, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.3334, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.14606741573033707, | |
| "grad_norm": 1.5196235179901123, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 0.3483, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15730337078651685, | |
| "grad_norm": 1.3063578605651855, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 0.318, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16853932584269662, | |
| "grad_norm": 1.255723237991333, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.3313, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1797752808988764, | |
| "grad_norm": 1.020233392715454, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 0.2873, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.19101123595505617, | |
| "grad_norm": 1.083106279373169, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 0.318, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.20224719101123595, | |
| "grad_norm": 0.828940749168396, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.2989, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.21348314606741572, | |
| "grad_norm": 0.7749581336975098, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 0.3038, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2247191011235955, | |
| "grad_norm": 0.7293257117271423, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.2886, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.23595505617977527, | |
| "grad_norm": 0.8117908835411072, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.3169, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.24719101123595505, | |
| "grad_norm": 0.7527502179145813, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 0.2727, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.25842696629213485, | |
| "grad_norm": 0.8296247124671936, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 0.2934, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2696629213483146, | |
| "grad_norm": 0.773894727230072, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.262, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.2808988764044944, | |
| "grad_norm": 0.8243966698646545, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.2648, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.29213483146067415, | |
| "grad_norm": 0.7368326783180237, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 0.3076, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.30337078651685395, | |
| "grad_norm": 0.7896155714988708, | |
| "learning_rate": 6e-06, | |
| "loss": 0.3105, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3146067415730337, | |
| "grad_norm": 0.7478722333908081, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 0.311, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3258426966292135, | |
| "grad_norm": 0.6260741949081421, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 0.2695, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.33707865168539325, | |
| "grad_norm": 0.7037319540977478, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.2642, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.34831460674157305, | |
| "grad_norm": 0.64321368932724, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 0.2843, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3595505617977528, | |
| "grad_norm": 0.699616014957428, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 0.2844, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3707865168539326, | |
| "grad_norm": 0.641768217086792, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.2978, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.38202247191011235, | |
| "grad_norm": 0.6910679936408997, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 0.2671, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.39325842696629215, | |
| "grad_norm": 0.5495407581329346, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 0.2497, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4044943820224719, | |
| "grad_norm": 0.5832553505897522, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.2556, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4157303370786517, | |
| "grad_norm": 0.5854374170303345, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 0.2596, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.42696629213483145, | |
| "grad_norm": 0.5847079753875732, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 0.2666, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.43820224719101125, | |
| "grad_norm": 0.5612176060676575, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.231, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.449438202247191, | |
| "grad_norm": 0.5835954546928406, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.2748, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4606741573033708, | |
| "grad_norm": 0.6874232888221741, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 0.294, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.47191011235955055, | |
| "grad_norm": 0.5141873955726624, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.2565, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.48314606741573035, | |
| "grad_norm": 0.5754961371421814, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 0.248, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4943820224719101, | |
| "grad_norm": 0.6071674823760986, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 0.2864, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5056179775280899, | |
| "grad_norm": 0.5854678153991699, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2873, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5168539325842697, | |
| "grad_norm": 0.5334392189979553, | |
| "learning_rate": 9.99984578822395e-06, | |
| "loss": 0.2508, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5280898876404494, | |
| "grad_norm": 0.6400332450866699, | |
| "learning_rate": 9.999383162408303e-06, | |
| "loss": 0.2412, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5393258426966292, | |
| "grad_norm": 0.6003167033195496, | |
| "learning_rate": 9.998612151090004e-06, | |
| "loss": 0.2551, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.550561797752809, | |
| "grad_norm": 0.5445740222930908, | |
| "learning_rate": 9.997532801828659e-06, | |
| "loss": 0.2519, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5617977528089888, | |
| "grad_norm": 0.563345730304718, | |
| "learning_rate": 9.996145181203616e-06, | |
| "loss": 0.2639, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5730337078651685, | |
| "grad_norm": 0.623471200466156, | |
| "learning_rate": 9.994449374809851e-06, | |
| "loss": 0.2351, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5842696629213483, | |
| "grad_norm": 0.5658097863197327, | |
| "learning_rate": 9.992445487252692e-06, | |
| "loss": 0.2528, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5955056179775281, | |
| "grad_norm": 0.5382084250450134, | |
| "learning_rate": 9.990133642141359e-06, | |
| "loss": 0.2433, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6067415730337079, | |
| "grad_norm": 0.5690349340438843, | |
| "learning_rate": 9.987513982081352e-06, | |
| "loss": 0.2658, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.6179775280898876, | |
| "grad_norm": 0.5317168235778809, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.2267, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6292134831460674, | |
| "grad_norm": 0.5717993974685669, | |
| "learning_rate": 9.981351882464706e-06, | |
| "loss": 0.2767, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6404494382022472, | |
| "grad_norm": 0.5172840356826782, | |
| "learning_rate": 9.9778098230154e-06, | |
| "loss": 0.2374, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.651685393258427, | |
| "grad_norm": 0.6190145611763, | |
| "learning_rate": 9.973960708808633e-06, | |
| "loss": 0.286, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6629213483146067, | |
| "grad_norm": 0.589586079120636, | |
| "learning_rate": 9.9698047772759e-06, | |
| "loss": 0.3017, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6741573033707865, | |
| "grad_norm": 0.4945083260536194, | |
| "learning_rate": 9.965342284774633e-06, | |
| "loss": 0.2324, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6853932584269663, | |
| "grad_norm": 0.5352405905723572, | |
| "learning_rate": 9.960573506572391e-06, | |
| "loss": 0.2714, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6966292134831461, | |
| "grad_norm": 0.5440102219581604, | |
| "learning_rate": 9.955498736829876e-06, | |
| "loss": 0.2758, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.7078651685393258, | |
| "grad_norm": 0.5334731340408325, | |
| "learning_rate": 9.95011828858279e-06, | |
| "loss": 0.2648, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.7191011235955056, | |
| "grad_norm": 0.506775975227356, | |
| "learning_rate": 9.944432493722525e-06, | |
| "loss": 0.2445, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.7303370786516854, | |
| "grad_norm": 0.5548930168151855, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.2597, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7415730337078652, | |
| "grad_norm": 0.5340548753738403, | |
| "learning_rate": 9.932146285882478e-06, | |
| "loss": 0.2499, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7528089887640449, | |
| "grad_norm": 0.6233317255973816, | |
| "learning_rate": 9.92554663077387e-06, | |
| "loss": 0.2964, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7640449438202247, | |
| "grad_norm": 0.5946013331413269, | |
| "learning_rate": 9.918643144747681e-06, | |
| "loss": 0.2892, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7752808988764045, | |
| "grad_norm": 0.5025975704193115, | |
| "learning_rate": 9.911436253643445e-06, | |
| "loss": 0.2569, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7865168539325843, | |
| "grad_norm": 0.5510412454605103, | |
| "learning_rate": 9.903926402016153e-06, | |
| "loss": 0.2298, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.797752808988764, | |
| "grad_norm": 0.5809914469718933, | |
| "learning_rate": 9.89611405310883e-06, | |
| "loss": 0.2564, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.8089887640449438, | |
| "grad_norm": 0.670417845249176, | |
| "learning_rate": 9.887999688823955e-06, | |
| "loss": 0.2906, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.8202247191011236, | |
| "grad_norm": 0.5335014462471008, | |
| "learning_rate": 9.879583809693737e-06, | |
| "loss": 0.2452, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.8314606741573034, | |
| "grad_norm": 0.5259139537811279, | |
| "learning_rate": 9.870866934849248e-06, | |
| "loss": 0.2807, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8426966292134831, | |
| "grad_norm": 0.5847296714782715, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.2384, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8539325842696629, | |
| "grad_norm": 0.549369752407074, | |
| "learning_rate": 9.852532367342712e-06, | |
| "loss": 0.245, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8651685393258427, | |
| "grad_norm": 0.5300970673561096, | |
| "learning_rate": 9.842915805643156e-06, | |
| "loss": 0.2581, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8764044943820225, | |
| "grad_norm": 0.5090511441230774, | |
| "learning_rate": 9.833000510084537e-06, | |
| "loss": 0.2508, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8876404494382022, | |
| "grad_norm": 0.5674286484718323, | |
| "learning_rate": 9.822787092288991e-06, | |
| "loss": 0.242, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.898876404494382, | |
| "grad_norm": 0.5166727900505066, | |
| "learning_rate": 9.812276182268236e-06, | |
| "loss": 0.2441, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9101123595505618, | |
| "grad_norm": 0.45713502168655396, | |
| "learning_rate": 9.801468428384716e-06, | |
| "loss": 0.2431, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.9213483146067416, | |
| "grad_norm": 0.5643267631530762, | |
| "learning_rate": 9.790364497311597e-06, | |
| "loss": 0.2433, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.9325842696629213, | |
| "grad_norm": 0.5387106537818909, | |
| "learning_rate": 9.778965073991652e-06, | |
| "loss": 0.2838, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9438202247191011, | |
| "grad_norm": 0.4815730154514313, | |
| "learning_rate": 9.767270861595006e-06, | |
| "loss": 0.2499, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9550561797752809, | |
| "grad_norm": 0.6593579053878784, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.2873, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9662921348314607, | |
| "grad_norm": 0.579657793045044, | |
| "learning_rate": 9.743000973127523e-06, | |
| "loss": 0.2539, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9775280898876404, | |
| "grad_norm": 0.5057744979858398, | |
| "learning_rate": 9.730426794137727e-06, | |
| "loss": 0.2628, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.9887640449438202, | |
| "grad_norm": 0.595583438873291, | |
| "learning_rate": 9.717560820140968e-06, | |
| "loss": 0.2391, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5546755194664001, | |
| "learning_rate": 9.704403844771128e-06, | |
| "loss": 0.2439, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.0112359550561798, | |
| "grad_norm": 0.6112611889839172, | |
| "learning_rate": 9.690956679612422e-06, | |
| "loss": 0.2126, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0224719101123596, | |
| "grad_norm": 0.5053119659423828, | |
| "learning_rate": 9.677220154149338e-06, | |
| "loss": 0.195, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.0337078651685394, | |
| "grad_norm": 0.48375797271728516, | |
| "learning_rate": 9.663195115715472e-06, | |
| "loss": 0.2014, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0449438202247192, | |
| "grad_norm": 0.4847674071788788, | |
| "learning_rate": 9.648882429441258e-06, | |
| "loss": 0.2242, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.0561797752808988, | |
| "grad_norm": 0.5149394869804382, | |
| "learning_rate": 9.634282978200605e-06, | |
| "loss": 0.2434, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.0674157303370786, | |
| "grad_norm": 0.5693023800849915, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.1967, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0786516853932584, | |
| "grad_norm": 0.5208962559700012, | |
| "learning_rate": 9.604227400705134e-06, | |
| "loss": 0.2029, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0898876404494382, | |
| "grad_norm": 0.46080583333969116, | |
| "learning_rate": 9.588773128419907e-06, | |
| "loss": 0.1977, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.101123595505618, | |
| "grad_norm": 0.46797895431518555, | |
| "learning_rate": 9.57303579899307e-06, | |
| "loss": 0.1958, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.1123595505617978, | |
| "grad_norm": 0.5413700938224792, | |
| "learning_rate": 9.557016383177226e-06, | |
| "loss": 0.1997, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.1235955056179776, | |
| "grad_norm": 0.5698537826538086, | |
| "learning_rate": 9.540715869125407e-06, | |
| "loss": 0.2097, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1348314606741572, | |
| "grad_norm": 0.5729172825813293, | |
| "learning_rate": 9.524135262330098e-06, | |
| "loss": 0.191, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.146067415730337, | |
| "grad_norm": 0.47994568943977356, | |
| "learning_rate": 9.507275585561228e-06, | |
| "loss": 0.2, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.1573033707865168, | |
| "grad_norm": 0.4811231195926666, | |
| "learning_rate": 9.490137878803078e-06, | |
| "loss": 0.2024, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.1685393258426966, | |
| "grad_norm": 0.5804319977760315, | |
| "learning_rate": 9.472723199190126e-06, | |
| "loss": 0.2213, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.1797752808988764, | |
| "grad_norm": 0.5076270699501038, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.1958, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1910112359550562, | |
| "grad_norm": 0.5329607725143433, | |
| "learning_rate": 9.437067235296418e-06, | |
| "loss": 0.1977, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.202247191011236, | |
| "grad_norm": 0.4917041063308716, | |
| "learning_rate": 9.418828150443469e-06, | |
| "loss": 0.2078, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.2134831460674158, | |
| "grad_norm": 0.4627450406551361, | |
| "learning_rate": 9.40031649145566e-06, | |
| "loss": 0.186, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.2247191011235956, | |
| "grad_norm": 0.4185917377471924, | |
| "learning_rate": 9.381533400219319e-06, | |
| "loss": 0.1792, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.2359550561797752, | |
| "grad_norm": 0.5121137499809265, | |
| "learning_rate": 9.362480035363987e-06, | |
| "loss": 0.2222, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.247191011235955, | |
| "grad_norm": 0.5793914198875427, | |
| "learning_rate": 9.343157572190957e-06, | |
| "loss": 0.2305, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.2584269662921348, | |
| "grad_norm": 0.5325943231582642, | |
| "learning_rate": 9.323567202600777e-06, | |
| "loss": 0.2148, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.2696629213483146, | |
| "grad_norm": 0.42034098505973816, | |
| "learning_rate": 9.30371013501972e-06, | |
| "loss": 0.178, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.2808988764044944, | |
| "grad_norm": 0.43560945987701416, | |
| "learning_rate": 9.28358759432525e-06, | |
| "loss": 0.1882, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.2921348314606742, | |
| "grad_norm": 0.4617180824279785, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.1825, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.303370786516854, | |
| "grad_norm": 0.5678491592407227, | |
| "learning_rate": 9.242551074907519e-06, | |
| "loss": 0.2219, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.3146067415730336, | |
| "grad_norm": 0.48582446575164795, | |
| "learning_rate": 9.221639627510076e-06, | |
| "loss": 0.1835, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.3258426966292136, | |
| "grad_norm": 0.44229453802108765, | |
| "learning_rate": 9.20046776949471e-06, | |
| "loss": 0.1699, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.3370786516853932, | |
| "grad_norm": 0.4907465875148773, | |
| "learning_rate": 9.179036806841352e-06, | |
| "loss": 0.2226, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.348314606741573, | |
| "grad_norm": 0.4677446782588959, | |
| "learning_rate": 9.157348061512728e-06, | |
| "loss": 0.1924, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3595505617977528, | |
| "grad_norm": 0.491872102022171, | |
| "learning_rate": 9.13540287137281e-06, | |
| "loss": 0.2151, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.3707865168539326, | |
| "grad_norm": 0.4544215798377991, | |
| "learning_rate": 9.1132025901043e-06, | |
| "loss": 0.1898, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.3820224719101124, | |
| "grad_norm": 0.4714052677154541, | |
| "learning_rate": 9.090748587125118e-06, | |
| "loss": 0.2279, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.3932584269662922, | |
| "grad_norm": 0.4766599237918854, | |
| "learning_rate": 9.068042247503937e-06, | |
| "loss": 0.2151, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.404494382022472, | |
| "grad_norm": 0.47084319591522217, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.1948, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.4157303370786516, | |
| "grad_norm": 0.4725306034088135, | |
| "learning_rate": 9.021878176350422e-06, | |
| "loss": 0.2025, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.4269662921348314, | |
| "grad_norm": 0.4789423942565918, | |
| "learning_rate": 8.998423292435455e-06, | |
| "loss": 0.1997, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.4382022471910112, | |
| "grad_norm": 0.4451747238636017, | |
| "learning_rate": 8.97472176693755e-06, | |
| "loss": 0.1687, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.449438202247191, | |
| "grad_norm": 0.5205036401748657, | |
| "learning_rate": 8.950775061878453e-06, | |
| "loss": 0.2318, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.4606741573033708, | |
| "grad_norm": 0.4217935800552368, | |
| "learning_rate": 8.926584654403725e-06, | |
| "loss": 0.1644, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4719101123595506, | |
| "grad_norm": 0.5485341548919678, | |
| "learning_rate": 8.902152036691649e-06, | |
| "loss": 0.209, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.4831460674157304, | |
| "grad_norm": 0.48224788904190063, | |
| "learning_rate": 8.877478715861173e-06, | |
| "loss": 0.2065, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.49438202247191, | |
| "grad_norm": 0.5133256316184998, | |
| "learning_rate": 8.852566213878947e-06, | |
| "loss": 0.1999, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.50561797752809, | |
| "grad_norm": 0.5534752011299133, | |
| "learning_rate": 8.827416067465442e-06, | |
| "loss": 0.2067, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.5168539325842696, | |
| "grad_norm": 0.49837929010391235, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.1921, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5280898876404494, | |
| "grad_norm": 0.45644959807395935, | |
| "learning_rate": 8.776409061425919e-06, | |
| "loss": 0.1966, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.5393258426966292, | |
| "grad_norm": 0.5083698630332947, | |
| "learning_rate": 8.750555348152299e-06, | |
| "loss": 0.2129, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.550561797752809, | |
| "grad_norm": 0.45993131399154663, | |
| "learning_rate": 8.72447028295811e-06, | |
| "loss": 0.2094, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.5617977528089888, | |
| "grad_norm": 0.5535938143730164, | |
| "learning_rate": 8.69815547489305e-06, | |
| "loss": 0.2375, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.5730337078651684, | |
| "grad_norm": 0.4734334349632263, | |
| "learning_rate": 8.671612547178428e-06, | |
| "loss": 0.1985, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5842696629213484, | |
| "grad_norm": 0.47312143445014954, | |
| "learning_rate": 8.644843137107058e-06, | |
| "loss": 0.1891, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.595505617977528, | |
| "grad_norm": 0.44169965386390686, | |
| "learning_rate": 8.617848895942246e-06, | |
| "loss": 0.189, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.606741573033708, | |
| "grad_norm": 0.48934826254844666, | |
| "learning_rate": 8.590631488815945e-06, | |
| "loss": 0.2177, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.6179775280898876, | |
| "grad_norm": 0.5859401822090149, | |
| "learning_rate": 8.563192594626027e-06, | |
| "loss": 0.2244, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.6292134831460674, | |
| "grad_norm": 0.5161330103874207, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.2016, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.6404494382022472, | |
| "grad_norm": 0.5324240922927856, | |
| "learning_rate": 8.50765712885428e-06, | |
| "loss": 0.2082, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.651685393258427, | |
| "grad_norm": 0.5002866983413696, | |
| "learning_rate": 8.479563982961572e-06, | |
| "loss": 0.2072, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.6629213483146068, | |
| "grad_norm": 0.5366979241371155, | |
| "learning_rate": 8.451256201172186e-06, | |
| "loss": 0.2255, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.6741573033707864, | |
| "grad_norm": 0.5217868685722351, | |
| "learning_rate": 8.422735529643445e-06, | |
| "loss": 0.1992, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.6853932584269664, | |
| "grad_norm": 0.4968906342983246, | |
| "learning_rate": 8.39400372766471e-06, | |
| "loss": 0.1907, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.696629213483146, | |
| "grad_norm": 0.46640124917030334, | |
| "learning_rate": 8.365062567548868e-06, | |
| "loss": 0.2149, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.7078651685393258, | |
| "grad_norm": 0.4526020288467407, | |
| "learning_rate": 8.335913834522999e-06, | |
| "loss": 0.1977, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.7191011235955056, | |
| "grad_norm": 0.5012973546981812, | |
| "learning_rate": 8.30655932661826e-06, | |
| "loss": 0.2005, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.7303370786516854, | |
| "grad_norm": 0.49397850036621094, | |
| "learning_rate": 8.27700085455897e-06, | |
| "loss": 0.2307, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.7415730337078652, | |
| "grad_norm": 0.4788162410259247, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.189, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7528089887640448, | |
| "grad_norm": 0.5018126368522644, | |
| "learning_rate": 8.217279323668895e-06, | |
| "loss": 0.2237, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.7640449438202248, | |
| "grad_norm": 0.5071457028388977, | |
| "learning_rate": 8.18711994874345e-06, | |
| "loss": 0.2058, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.7752808988764044, | |
| "grad_norm": 0.4464392066001892, | |
| "learning_rate": 8.15676397724689e-06, | |
| "loss": 0.2093, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.7865168539325844, | |
| "grad_norm": 0.6191111207008362, | |
| "learning_rate": 8.126213281678527e-06, | |
| "loss": 0.2286, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.797752808988764, | |
| "grad_norm": 0.4710521399974823, | |
| "learning_rate": 8.095469746549172e-06, | |
| "loss": 0.2043, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.8089887640449438, | |
| "grad_norm": 0.47517672181129456, | |
| "learning_rate": 8.064535268264883e-06, | |
| "loss": 0.2182, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.8202247191011236, | |
| "grad_norm": 0.48008614778518677, | |
| "learning_rate": 8.033411755009999e-06, | |
| "loss": 0.2221, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.8314606741573034, | |
| "grad_norm": 0.4443986415863037, | |
| "learning_rate": 8.002101126629422e-06, | |
| "loss": 0.1951, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.8426966292134832, | |
| "grad_norm": 0.4856914281845093, | |
| "learning_rate": 7.970605314510194e-06, | |
| "loss": 0.1888, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.8539325842696628, | |
| "grad_norm": 0.4712364077568054, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.1907, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.8651685393258428, | |
| "grad_norm": 0.486104279756546, | |
| "learning_rate": 7.907065921599153e-06, | |
| "loss": 0.2124, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.8764044943820224, | |
| "grad_norm": 0.5030559301376343, | |
| "learning_rate": 7.875026260216395e-06, | |
| "loss": 0.2129, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.8876404494382022, | |
| "grad_norm": 0.5313302278518677, | |
| "learning_rate": 7.842809253671321e-06, | |
| "loss": 0.2208, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.898876404494382, | |
| "grad_norm": 0.4822382926940918, | |
| "learning_rate": 7.810416889260653e-06, | |
| "loss": 0.2106, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.9101123595505618, | |
| "grad_norm": 0.5109752416610718, | |
| "learning_rate": 7.777851165098012e-06, | |
| "loss": 0.2203, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.9213483146067416, | |
| "grad_norm": 0.49171018600463867, | |
| "learning_rate": 7.74511408999066e-06, | |
| "loss": 0.2057, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.9325842696629212, | |
| "grad_norm": 0.4889599680900574, | |
| "learning_rate": 7.712207683315595e-06, | |
| "loss": 0.1964, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.9438202247191012, | |
| "grad_norm": 0.48753100633621216, | |
| "learning_rate": 7.679133974894984e-06, | |
| "loss": 0.2192, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.9550561797752808, | |
| "grad_norm": 0.47565633058547974, | |
| "learning_rate": 7.645895004870953e-06, | |
| "loss": 0.2104, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.9662921348314608, | |
| "grad_norm": 0.49253159761428833, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.1932, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9775280898876404, | |
| "grad_norm": 0.46585404872894287, | |
| "learning_rate": 7.5789294914252376e-06, | |
| "loss": 0.2148, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.9887640449438202, | |
| "grad_norm": 0.4452122449874878, | |
| "learning_rate": 7.545207078751858e-06, | |
| "loss": 0.1978, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.503316342830658, | |
| "learning_rate": 7.511327665716863e-06, | |
| "loss": 0.1975, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.0112359550561796, | |
| "grad_norm": 0.5127564668655396, | |
| "learning_rate": 7.477293342162038e-06, | |
| "loss": 0.1344, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.0224719101123596, | |
| "grad_norm": 0.46806156635284424, | |
| "learning_rate": 7.443106207484776e-06, | |
| "loss": 0.1438, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.033707865168539, | |
| "grad_norm": 0.4583097994327545, | |
| "learning_rate": 7.408768370508577e-06, | |
| "loss": 0.1383, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.044943820224719, | |
| "grad_norm": 0.41952410340309143, | |
| "learning_rate": 7.3742819493529725e-06, | |
| "loss": 0.1247, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.056179775280899, | |
| "grad_norm": 0.6133317947387695, | |
| "learning_rate": 7.3396490713028674e-06, | |
| "loss": 0.1415, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.067415730337079, | |
| "grad_norm": 0.5251442193984985, | |
| "learning_rate": 7.304871872677313e-06, | |
| "loss": 0.1214, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.0786516853932584, | |
| "grad_norm": 0.4410766065120697, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.1281, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.0898876404494384, | |
| "grad_norm": 0.4335516393184662, | |
| "learning_rate": 7.2348931033556065e-06, | |
| "loss": 0.1389, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.101123595505618, | |
| "grad_norm": 0.6064698100090027, | |
| "learning_rate": 7.199695849279576e-06, | |
| "loss": 0.1848, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.1123595505617976, | |
| "grad_norm": 0.45688655972480774, | |
| "learning_rate": 7.164362907602072e-06, | |
| "loss": 0.1286, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.1235955056179776, | |
| "grad_norm": 0.4466540515422821, | |
| "learning_rate": 7.128896457825364e-06, | |
| "loss": 0.1317, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.134831460674157, | |
| "grad_norm": 0.4052949547767639, | |
| "learning_rate": 7.093298687687141e-06, | |
| "loss": 0.1149, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.146067415730337, | |
| "grad_norm": 0.48267531394958496, | |
| "learning_rate": 7.057571793025545e-06, | |
| "loss": 0.1427, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.157303370786517, | |
| "grad_norm": 0.4643622934818268, | |
| "learning_rate": 7.021717977643726e-06, | |
| "loss": 0.1324, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.168539325842697, | |
| "grad_norm": 0.5360526442527771, | |
| "learning_rate": 6.985739453173903e-06, | |
| "loss": 0.1501, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.1797752808988764, | |
| "grad_norm": 0.4406614899635315, | |
| "learning_rate": 6.949638438940942e-06, | |
| "loss": 0.1331, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.191011235955056, | |
| "grad_norm": 0.46032071113586426, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.1367, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.202247191011236, | |
| "grad_norm": 0.4472227990627289, | |
| "learning_rate": 6.877077856126416e-06, | |
| "loss": 0.1562, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.2134831460674156, | |
| "grad_norm": 0.45712995529174805, | |
| "learning_rate": 6.840622763423391e-06, | |
| "loss": 0.1276, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.2247191011235956, | |
| "grad_norm": 0.43871307373046875, | |
| "learning_rate": 6.804054132438209e-06, | |
| "loss": 0.1413, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.235955056179775, | |
| "grad_norm": 0.43786612153053284, | |
| "learning_rate": 6.767374218896286e-06, | |
| "loss": 0.1405, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.247191011235955, | |
| "grad_norm": 0.4727621376514435, | |
| "learning_rate": 6.730585285387465e-06, | |
| "loss": 0.1415, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.258426966292135, | |
| "grad_norm": 0.4680100679397583, | |
| "learning_rate": 6.693689601226458e-06, | |
| "loss": 0.1238, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.2696629213483144, | |
| "grad_norm": 0.4155004918575287, | |
| "learning_rate": 6.656689442312855e-06, | |
| "loss": 0.133, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.2808988764044944, | |
| "grad_norm": 0.45196136832237244, | |
| "learning_rate": 6.619587090990748e-06, | |
| "loss": 0.1355, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.292134831460674, | |
| "grad_norm": 0.3976186513900757, | |
| "learning_rate": 6.582384835907931e-06, | |
| "loss": 0.1155, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.303370786516854, | |
| "grad_norm": 0.39685821533203125, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.1413, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.3146067415730336, | |
| "grad_norm": 0.44746285676956177, | |
| "learning_rate": 6.507689799722479e-06, | |
| "loss": 0.1323, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.3258426966292136, | |
| "grad_norm": 0.45192277431488037, | |
| "learning_rate": 6.47020162616152e-06, | |
| "loss": 0.1244, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.337078651685393, | |
| "grad_norm": 0.367249071598053, | |
| "learning_rate": 6.432622763638993e-06, | |
| "loss": 0.1111, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.348314606741573, | |
| "grad_norm": 0.42031487822532654, | |
| "learning_rate": 6.3949555301961474e-06, | |
| "loss": 0.13, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.359550561797753, | |
| "grad_norm": 0.4596504271030426, | |
| "learning_rate": 6.3572022493253715e-06, | |
| "loss": 0.1328, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.370786516853933, | |
| "grad_norm": 0.45772069692611694, | |
| "learning_rate": 6.3193652498268656e-06, | |
| "loss": 0.136, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.3820224719101124, | |
| "grad_norm": 0.40773630142211914, | |
| "learning_rate": 6.281446865664984e-06, | |
| "loss": 0.1124, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.393258426966292, | |
| "grad_norm": 0.5169959664344788, | |
| "learning_rate": 6.243449435824276e-06, | |
| "loss": 0.1576, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.404494382022472, | |
| "grad_norm": 0.41499051451683044, | |
| "learning_rate": 6.205375304165194e-06, | |
| "loss": 0.1261, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.4157303370786516, | |
| "grad_norm": 0.41324612498283386, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.1175, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.4269662921348316, | |
| "grad_norm": 0.467218816280365, | |
| "learning_rate": 6.1290063343455196e-06, | |
| "loss": 0.14, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.438202247191011, | |
| "grad_norm": 0.41227272152900696, | |
| "learning_rate": 6.090716206982714e-06, | |
| "loss": 0.1274, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.449438202247191, | |
| "grad_norm": 0.40290334820747375, | |
| "learning_rate": 6.052358799106528e-06, | |
| "loss": 0.1429, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.460674157303371, | |
| "grad_norm": 0.41341057419776917, | |
| "learning_rate": 6.013936476782563e-06, | |
| "loss": 0.1166, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.4719101123595504, | |
| "grad_norm": 0.4011130630970001, | |
| "learning_rate": 5.975451610080643e-06, | |
| "loss": 0.1278, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.4831460674157304, | |
| "grad_norm": 0.47281768918037415, | |
| "learning_rate": 5.936906572928625e-06, | |
| "loss": 0.1359, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.49438202247191, | |
| "grad_norm": 0.4146575629711151, | |
| "learning_rate": 5.898303742965964e-06, | |
| "loss": 0.1427, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.50561797752809, | |
| "grad_norm": 0.4124404191970825, | |
| "learning_rate": 5.859645501397048e-06, | |
| "loss": 0.1274, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.5168539325842696, | |
| "grad_norm": 0.3924229145050049, | |
| "learning_rate": 5.820934232844315e-06, | |
| "loss": 0.1296, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.5280898876404496, | |
| "grad_norm": 0.40923216938972473, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.1206, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.539325842696629, | |
| "grad_norm": 0.4435917139053345, | |
| "learning_rate": 5.743362169484617e-06, | |
| "loss": 0.1227, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.550561797752809, | |
| "grad_norm": 0.4076763689517975, | |
| "learning_rate": 5.704506159687914e-06, | |
| "loss": 0.1237, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.561797752808989, | |
| "grad_norm": 0.4328089654445648, | |
| "learning_rate": 5.665606692632762e-06, | |
| "loss": 0.1429, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.5730337078651684, | |
| "grad_norm": 0.4347201883792877, | |
| "learning_rate": 5.626666167821522e-06, | |
| "loss": 0.1389, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.5842696629213484, | |
| "grad_norm": 0.4432125389575958, | |
| "learning_rate": 5.587686987289189e-06, | |
| "loss": 0.1348, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.595505617977528, | |
| "grad_norm": 0.4269976317882538, | |
| "learning_rate": 5.548671555455226e-06, | |
| "loss": 0.1288, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.606741573033708, | |
| "grad_norm": 0.4509507119655609, | |
| "learning_rate": 5.50962227897525e-06, | |
| "loss": 0.1439, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.6179775280898876, | |
| "grad_norm": 0.48018571734428406, | |
| "learning_rate": 5.470541566592573e-06, | |
| "loss": 0.1472, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.629213483146067, | |
| "grad_norm": 0.42168352007865906, | |
| "learning_rate": 5.431431828989618e-06, | |
| "loss": 0.1365, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.640449438202247, | |
| "grad_norm": 0.4211483895778656, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.1362, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.6516853932584272, | |
| "grad_norm": 0.39289364218711853, | |
| "learning_rate": 5.353134929655834e-06, | |
| "loss": 0.1283, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.662921348314607, | |
| "grad_norm": 0.4453902840614319, | |
| "learning_rate": 5.3139525976465675e-06, | |
| "loss": 0.1359, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.6741573033707864, | |
| "grad_norm": 0.4308948516845703, | |
| "learning_rate": 5.27475089956223e-06, | |
| "loss": 0.1349, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.6853932584269664, | |
| "grad_norm": 0.41243335604667664, | |
| "learning_rate": 5.235532253548213e-06, | |
| "loss": 0.133, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.696629213483146, | |
| "grad_norm": 0.4549100995063782, | |
| "learning_rate": 5.1962990787953436e-06, | |
| "loss": 0.1345, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.7078651685393256, | |
| "grad_norm": 0.40238845348358154, | |
| "learning_rate": 5.157053795390642e-06, | |
| "loss": 0.1124, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.7191011235955056, | |
| "grad_norm": 0.4336741864681244, | |
| "learning_rate": 5.117798824168052e-06, | |
| "loss": 0.1291, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.7303370786516856, | |
| "grad_norm": 0.4185507595539093, | |
| "learning_rate": 5.078536586559104e-06, | |
| "loss": 0.1446, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.741573033707865, | |
| "grad_norm": 0.4253235161304474, | |
| "learning_rate": 5.039269504443557e-06, | |
| "loss": 0.1446, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.752808988764045, | |
| "grad_norm": 0.4910286068916321, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1451, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.764044943820225, | |
| "grad_norm": 0.4122835397720337, | |
| "learning_rate": 4.9607304955564456e-06, | |
| "loss": 0.1125, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.7752808988764044, | |
| "grad_norm": 0.41325777769088745, | |
| "learning_rate": 4.921463413440898e-06, | |
| "loss": 0.1324, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.7865168539325844, | |
| "grad_norm": 0.43836483359336853, | |
| "learning_rate": 4.88220117583195e-06, | |
| "loss": 0.1474, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.797752808988764, | |
| "grad_norm": 0.3876822590827942, | |
| "learning_rate": 4.842946204609359e-06, | |
| "loss": 0.1362, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.808988764044944, | |
| "grad_norm": 0.44368717074394226, | |
| "learning_rate": 4.803700921204659e-06, | |
| "loss": 0.1511, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.8202247191011236, | |
| "grad_norm": 0.43904393911361694, | |
| "learning_rate": 4.7644677464517874e-06, | |
| "loss": 0.1285, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.831460674157303, | |
| "grad_norm": 0.43354493379592896, | |
| "learning_rate": 4.725249100437773e-06, | |
| "loss": 0.1237, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.842696629213483, | |
| "grad_norm": 0.4428333044052124, | |
| "learning_rate": 4.686047402353433e-06, | |
| "loss": 0.1338, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.853932584269663, | |
| "grad_norm": 0.4285905063152313, | |
| "learning_rate": 4.646865070344168e-06, | |
| "loss": 0.1382, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.865168539325843, | |
| "grad_norm": 0.44894492626190186, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.14, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.8764044943820224, | |
| "grad_norm": 0.4365105628967285, | |
| "learning_rate": 4.568568171010384e-06, | |
| "loss": 0.1254, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.8876404494382024, | |
| "grad_norm": 0.4431806802749634, | |
| "learning_rate": 4.529458433407429e-06, | |
| "loss": 0.1389, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.898876404494382, | |
| "grad_norm": 0.4079169034957886, | |
| "learning_rate": 4.490377721024751e-06, | |
| "loss": 0.1129, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.9101123595505616, | |
| "grad_norm": 0.38153865933418274, | |
| "learning_rate": 4.451328444544774e-06, | |
| "loss": 0.1341, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.9213483146067416, | |
| "grad_norm": 0.4263458847999573, | |
| "learning_rate": 4.4123130127108125e-06, | |
| "loss": 0.146, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.932584269662921, | |
| "grad_norm": 0.41687625646591187, | |
| "learning_rate": 4.373333832178478e-06, | |
| "loss": 0.1394, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.943820224719101, | |
| "grad_norm": 0.40404847264289856, | |
| "learning_rate": 4.3343933073672395e-06, | |
| "loss": 0.1179, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.955056179775281, | |
| "grad_norm": 0.3679511547088623, | |
| "learning_rate": 4.295493840312087e-06, | |
| "loss": 0.1073, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.966292134831461, | |
| "grad_norm": 0.46775326132774353, | |
| "learning_rate": 4.256637830515385e-06, | |
| "loss": 0.1409, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.9775280898876404, | |
| "grad_norm": 0.4219604730606079, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.1252, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.98876404494382, | |
| "grad_norm": 0.4047105610370636, | |
| "learning_rate": 4.179065767155686e-06, | |
| "loss": 0.1349, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.4388047456741333, | |
| "learning_rate": 4.140354498602952e-06, | |
| "loss": 0.1283, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.0112359550561796, | |
| "grad_norm": 0.5455737709999084, | |
| "learning_rate": 4.1016962570340375e-06, | |
| "loss": 0.0822, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.0224719101123596, | |
| "grad_norm": 0.5572327375411987, | |
| "learning_rate": 4.063093427071376e-06, | |
| "loss": 0.0927, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.033707865168539, | |
| "grad_norm": 0.46223390102386475, | |
| "learning_rate": 4.02454838991936e-06, | |
| "loss": 0.0884, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.044943820224719, | |
| "grad_norm": 0.4062277376651764, | |
| "learning_rate": 3.986063523217439e-06, | |
| "loss": 0.0819, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.056179775280899, | |
| "grad_norm": 0.4413997232913971, | |
| "learning_rate": 3.947641200893473e-06, | |
| "loss": 0.0712, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.067415730337079, | |
| "grad_norm": 0.7831788659095764, | |
| "learning_rate": 3.909283793017289e-06, | |
| "loss": 0.0899, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.0786516853932584, | |
| "grad_norm": 0.6694844365119934, | |
| "learning_rate": 3.870993665654482e-06, | |
| "loss": 0.0899, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.0898876404494384, | |
| "grad_norm": 0.44210654497146606, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.0809, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.101123595505618, | |
| "grad_norm": 0.37950965762138367, | |
| "learning_rate": 3.7946246958348077e-06, | |
| "loss": 0.0742, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.1123595505617976, | |
| "grad_norm": 0.3797522783279419, | |
| "learning_rate": 3.756550564175727e-06, | |
| "loss": 0.0689, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.1235955056179776, | |
| "grad_norm": 0.39495426416397095, | |
| "learning_rate": 3.7185531343350167e-06, | |
| "loss": 0.0724, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.134831460674157, | |
| "grad_norm": 0.466620534658432, | |
| "learning_rate": 3.680634750173137e-06, | |
| "loss": 0.0867, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.146067415730337, | |
| "grad_norm": 0.445311576128006, | |
| "learning_rate": 3.6427977506746293e-06, | |
| "loss": 0.0705, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.157303370786517, | |
| "grad_norm": 0.41297537088394165, | |
| "learning_rate": 3.6050444698038547e-06, | |
| "loss": 0.0852, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.168539325842697, | |
| "grad_norm": 0.3536671996116638, | |
| "learning_rate": 3.5673772363610083e-06, | |
| "loss": 0.0639, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.1797752808988764, | |
| "grad_norm": 0.39393407106399536, | |
| "learning_rate": 3.5297983738384813e-06, | |
| "loss": 0.0753, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.191011235955056, | |
| "grad_norm": 0.4105747938156128, | |
| "learning_rate": 3.492310200277522e-06, | |
| "loss": 0.0704, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.202247191011236, | |
| "grad_norm": 0.4793269634246826, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.0757, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.2134831460674156, | |
| "grad_norm": 0.4588974118232727, | |
| "learning_rate": 3.4176151640920696e-06, | |
| "loss": 0.0869, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.2247191011235956, | |
| "grad_norm": 0.48513558506965637, | |
| "learning_rate": 3.3804129090092542e-06, | |
| "loss": 0.078, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.235955056179775, | |
| "grad_norm": 0.38135215640068054, | |
| "learning_rate": 3.3433105576871448e-06, | |
| "loss": 0.0754, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.247191011235955, | |
| "grad_norm": 0.40388548374176025, | |
| "learning_rate": 3.3063103987735433e-06, | |
| "loss": 0.0713, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.258426966292135, | |
| "grad_norm": 0.3911704123020172, | |
| "learning_rate": 3.269414714612534e-06, | |
| "loss": 0.0761, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.2696629213483144, | |
| "grad_norm": 0.4082418978214264, | |
| "learning_rate": 3.2326257811037154e-06, | |
| "loss": 0.0724, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.2808988764044944, | |
| "grad_norm": 0.4108784794807434, | |
| "learning_rate": 3.195945867561791e-06, | |
| "loss": 0.0759, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.292134831460674, | |
| "grad_norm": 0.3417165279388428, | |
| "learning_rate": 3.1593772365766107e-06, | |
| "loss": 0.0661, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.303370786516854, | |
| "grad_norm": 0.3870854675769806, | |
| "learning_rate": 3.122922143873584e-06, | |
| "loss": 0.0797, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.3146067415730336, | |
| "grad_norm": 0.4093579947948456, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.0774, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.3258426966292136, | |
| "grad_norm": 0.3988387882709503, | |
| "learning_rate": 3.0503615610590605e-06, | |
| "loss": 0.084, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.337078651685393, | |
| "grad_norm": 0.3742392957210541, | |
| "learning_rate": 3.0142605468260976e-06, | |
| "loss": 0.0775, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.348314606741573, | |
| "grad_norm": 0.39700236916542053, | |
| "learning_rate": 2.9782820223562758e-06, | |
| "loss": 0.0721, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.359550561797753, | |
| "grad_norm": 0.4066769778728485, | |
| "learning_rate": 2.9424282069744564e-06, | |
| "loss": 0.0809, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.370786516853933, | |
| "grad_norm": 0.37812596559524536, | |
| "learning_rate": 2.906701312312861e-06, | |
| "loss": 0.0763, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.3820224719101124, | |
| "grad_norm": 0.4282001852989197, | |
| "learning_rate": 2.871103542174637e-06, | |
| "loss": 0.0769, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.393258426966292, | |
| "grad_norm": 0.4477894604206085, | |
| "learning_rate": 2.8356370923979326e-06, | |
| "loss": 0.0859, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.404494382022472, | |
| "grad_norm": 0.36554381251335144, | |
| "learning_rate": 2.800304150720424e-06, | |
| "loss": 0.075, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.4157303370786516, | |
| "grad_norm": 0.398842990398407, | |
| "learning_rate": 2.765106896644395e-06, | |
| "loss": 0.0832, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.4269662921348316, | |
| "grad_norm": 0.36652693152427673, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.0661, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.438202247191011, | |
| "grad_norm": 0.3799467384815216, | |
| "learning_rate": 2.6951281273226894e-06, | |
| "loss": 0.0756, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.449438202247191, | |
| "grad_norm": 0.39734625816345215, | |
| "learning_rate": 2.6603509286971342e-06, | |
| "loss": 0.0795, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.460674157303371, | |
| "grad_norm": 0.36823272705078125, | |
| "learning_rate": 2.6257180506470283e-06, | |
| "loss": 0.0771, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.4719101123595504, | |
| "grad_norm": 0.4019593298435211, | |
| "learning_rate": 2.5912316294914232e-06, | |
| "loss": 0.0847, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.4831460674157304, | |
| "grad_norm": 0.3722269833087921, | |
| "learning_rate": 2.5568937925152272e-06, | |
| "loss": 0.0786, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.49438202247191, | |
| "grad_norm": 0.4078711271286011, | |
| "learning_rate": 2.5227066578379624e-06, | |
| "loss": 0.0872, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.50561797752809, | |
| "grad_norm": 0.3710861802101135, | |
| "learning_rate": 2.4886723342831375e-06, | |
| "loss": 0.0863, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 3.5168539325842696, | |
| "grad_norm": 0.36387649178504944, | |
| "learning_rate": 2.4547929212481436e-06, | |
| "loss": 0.0741, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 3.5280898876404496, | |
| "grad_norm": 0.33584874868392944, | |
| "learning_rate": 2.4210705085747633e-06, | |
| "loss": 0.0808, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.539325842696629, | |
| "grad_norm": 0.3632005453109741, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.0639, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 3.550561797752809, | |
| "grad_norm": 0.3687782883644104, | |
| "learning_rate": 2.354104995129048e-06, | |
| "loss": 0.0746, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.561797752808989, | |
| "grad_norm": 0.3989294469356537, | |
| "learning_rate": 2.320866025105016e-06, | |
| "loss": 0.0653, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 3.5730337078651684, | |
| "grad_norm": 0.33881452679634094, | |
| "learning_rate": 2.2877923166844073e-06, | |
| "loss": 0.0732, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.5842696629213484, | |
| "grad_norm": 0.3820963203907013, | |
| "learning_rate": 2.254885910009341e-06, | |
| "loss": 0.0653, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 3.595505617977528, | |
| "grad_norm": 0.42072489857673645, | |
| "learning_rate": 2.2221488349019903e-06, | |
| "loss": 0.0815, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.606741573033708, | |
| "grad_norm": 0.3743472993373871, | |
| "learning_rate": 2.1895831107393485e-06, | |
| "loss": 0.076, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 3.6179775280898876, | |
| "grad_norm": 0.3808266818523407, | |
| "learning_rate": 2.15719074632868e-06, | |
| "loss": 0.0705, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.629213483146067, | |
| "grad_norm": 0.4201536774635315, | |
| "learning_rate": 2.124973739783609e-06, | |
| "loss": 0.0859, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 3.640449438202247, | |
| "grad_norm": 0.38096362352371216, | |
| "learning_rate": 2.0929340784008474e-06, | |
| "loss": 0.074, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.6516853932584272, | |
| "grad_norm": 0.39911383390426636, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.0805, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.662921348314607, | |
| "grad_norm": 0.32940348982810974, | |
| "learning_rate": 2.029394685489808e-06, | |
| "loss": 0.0616, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.6741573033707864, | |
| "grad_norm": 0.3778252601623535, | |
| "learning_rate": 1.9978988733705807e-06, | |
| "loss": 0.0872, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 3.6853932584269664, | |
| "grad_norm": 0.3902676999568939, | |
| "learning_rate": 1.9665882449900024e-06, | |
| "loss": 0.0656, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.696629213483146, | |
| "grad_norm": 0.3568686544895172, | |
| "learning_rate": 1.9354647317351187e-06, | |
| "loss": 0.0631, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 3.7078651685393256, | |
| "grad_norm": 0.38821014761924744, | |
| "learning_rate": 1.9045302534508298e-06, | |
| "loss": 0.0737, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.7191011235955056, | |
| "grad_norm": 0.3225777745246887, | |
| "learning_rate": 1.873786718321476e-06, | |
| "loss": 0.0652, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 3.7303370786516856, | |
| "grad_norm": 0.33293595910072327, | |
| "learning_rate": 1.8432360227531116e-06, | |
| "loss": 0.0667, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.741573033707865, | |
| "grad_norm": 0.38181042671203613, | |
| "learning_rate": 1.8128800512565514e-06, | |
| "loss": 0.0766, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.752808988764045, | |
| "grad_norm": 0.35045260190963745, | |
| "learning_rate": 1.7827206763311055e-06, | |
| "loss": 0.0721, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.764044943820225, | |
| "grad_norm": 0.3418102562427521, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.0818, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.7752808988764044, | |
| "grad_norm": 0.33109530806541443, | |
| "learning_rate": 1.722999145441031e-06, | |
| "loss": 0.064, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.7865168539325844, | |
| "grad_norm": 0.37899285554885864, | |
| "learning_rate": 1.6934406733817417e-06, | |
| "loss": 0.0739, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.797752808988764, | |
| "grad_norm": 0.3409174084663391, | |
| "learning_rate": 1.6640861654770007e-06, | |
| "loss": 0.0666, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.808988764044944, | |
| "grad_norm": 0.3983001708984375, | |
| "learning_rate": 1.6349374324511347e-06, | |
| "loss": 0.0817, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.8202247191011236, | |
| "grad_norm": 0.36114805936813354, | |
| "learning_rate": 1.6059962723352912e-06, | |
| "loss": 0.074, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.831460674157303, | |
| "grad_norm": 0.39180055260658264, | |
| "learning_rate": 1.5772644703565564e-06, | |
| "loss": 0.0776, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.842696629213483, | |
| "grad_norm": 0.37725961208343506, | |
| "learning_rate": 1.5487437988278141e-06, | |
| "loss": 0.0741, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.853932584269663, | |
| "grad_norm": 0.3951829671859741, | |
| "learning_rate": 1.5204360170384286e-06, | |
| "loss": 0.0751, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.865168539325843, | |
| "grad_norm": 0.41868656873703003, | |
| "learning_rate": 1.4923428711457217e-06, | |
| "loss": 0.0727, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.8764044943820224, | |
| "grad_norm": 0.35519301891326904, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.0738, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.8876404494382024, | |
| "grad_norm": 0.41053545475006104, | |
| "learning_rate": 1.4368074053739733e-06, | |
| "loss": 0.089, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.898876404494382, | |
| "grad_norm": 0.4047039747238159, | |
| "learning_rate": 1.4093685111840567e-06, | |
| "loss": 0.0659, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.9101123595505616, | |
| "grad_norm": 0.37193068861961365, | |
| "learning_rate": 1.382151104057754e-06, | |
| "loss": 0.0718, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.9213483146067416, | |
| "grad_norm": 0.3433876037597656, | |
| "learning_rate": 1.3551568628929434e-06, | |
| "loss": 0.0608, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.932584269662921, | |
| "grad_norm": 0.3564971685409546, | |
| "learning_rate": 1.3283874528215735e-06, | |
| "loss": 0.0713, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.943820224719101, | |
| "grad_norm": 0.3533005118370056, | |
| "learning_rate": 1.301844525106951e-06, | |
| "loss": 0.0722, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.955056179775281, | |
| "grad_norm": 0.4084896147251129, | |
| "learning_rate": 1.2755297170418913e-06, | |
| "loss": 0.0759, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.966292134831461, | |
| "grad_norm": 0.3923269510269165, | |
| "learning_rate": 1.2494446518477022e-06, | |
| "loss": 0.0907, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.9775280898876404, | |
| "grad_norm": 0.3590816259384155, | |
| "learning_rate": 1.2235909385740825e-06, | |
| "loss": 0.0785, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.98876404494382, | |
| "grad_norm": 0.4006705582141876, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.0826, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.3407679498195648, | |
| "learning_rate": 1.1725839325345601e-06, | |
| "loss": 0.0709, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.01123595505618, | |
| "grad_norm": 0.4902256727218628, | |
| "learning_rate": 1.1474337861210543e-06, | |
| "loss": 0.0404, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.022471910112359, | |
| "grad_norm": 0.4681622087955475, | |
| "learning_rate": 1.1225212841388282e-06, | |
| "loss": 0.0441, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.033707865168539, | |
| "grad_norm": 0.47430193424224854, | |
| "learning_rate": 1.097847963308351e-06, | |
| "loss": 0.0503, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.044943820224719, | |
| "grad_norm": 0.4806967079639435, | |
| "learning_rate": 1.0734153455962765e-06, | |
| "loss": 0.0491, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.056179775280899, | |
| "grad_norm": 0.4137580990791321, | |
| "learning_rate": 1.049224938121548e-06, | |
| "loss": 0.0501, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.067415730337078, | |
| "grad_norm": 0.36525774002075195, | |
| "learning_rate": 1.02527823306245e-06, | |
| "loss": 0.0395, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.078651685393258, | |
| "grad_norm": 0.3174494206905365, | |
| "learning_rate": 1.0015767075645472e-06, | |
| "loss": 0.0328, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.089887640449438, | |
| "grad_norm": 0.31036630272865295, | |
| "learning_rate": 9.781218236495776e-07, | |
| "loss": 0.042, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.101123595505618, | |
| "grad_norm": 0.32368966937065125, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.0498, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.112359550561798, | |
| "grad_norm": 0.3594004511833191, | |
| "learning_rate": 9.319577524960655e-07, | |
| "loss": 0.0477, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.123595505617978, | |
| "grad_norm": 0.3772434890270233, | |
| "learning_rate": 9.09251412874882e-07, | |
| "loss": 0.0379, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.134831460674158, | |
| "grad_norm": 0.47112375497817993, | |
| "learning_rate": 8.867974098957016e-07, | |
| "loss": 0.0385, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.146067415730337, | |
| "grad_norm": 0.41787606477737427, | |
| "learning_rate": 8.645971286271903e-07, | |
| "loss": 0.0377, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.157303370786517, | |
| "grad_norm": 0.39339321851730347, | |
| "learning_rate": 8.426519384872733e-07, | |
| "loss": 0.0384, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.168539325842697, | |
| "grad_norm": 0.37326088547706604, | |
| "learning_rate": 8.209631931586499e-07, | |
| "loss": 0.044, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.179775280898877, | |
| "grad_norm": 0.3865525424480438, | |
| "learning_rate": 7.995322305052905e-07, | |
| "loss": 0.0501, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.191011235955056, | |
| "grad_norm": 0.3712690770626068, | |
| "learning_rate": 7.783603724899258e-07, | |
| "loss": 0.0402, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.202247191011236, | |
| "grad_norm": 0.3322262465953827, | |
| "learning_rate": 7.574489250924821e-07, | |
| "loss": 0.037, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.213483146067416, | |
| "grad_norm": 0.32212916016578674, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.0369, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.224719101123595, | |
| "grad_norm": 0.32061728835105896, | |
| "learning_rate": 7.164124056747523e-07, | |
| "loss": 0.0342, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.235955056179775, | |
| "grad_norm": 0.303558349609375, | |
| "learning_rate": 6.962898649802824e-07, | |
| "loss": 0.0418, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.247191011235955, | |
| "grad_norm": 0.32052311301231384, | |
| "learning_rate": 6.764327973992252e-07, | |
| "loss": 0.0442, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.258426966292135, | |
| "grad_norm": 0.31100547313690186, | |
| "learning_rate": 6.568424278090446e-07, | |
| "loss": 0.0385, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.269662921348314, | |
| "grad_norm": 0.32819902896881104, | |
| "learning_rate": 6.375199646360142e-07, | |
| "loss": 0.0506, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.280898876404494, | |
| "grad_norm": 0.33846989274024963, | |
| "learning_rate": 6.184665997806832e-07, | |
| "loss": 0.0348, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.292134831460674, | |
| "grad_norm": 0.31254467368125916, | |
| "learning_rate": 5.996835085443403e-07, | |
| "loss": 0.0421, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.303370786516854, | |
| "grad_norm": 0.3004555106163025, | |
| "learning_rate": 5.811718495565327e-07, | |
| "loss": 0.036, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.314606741573034, | |
| "grad_norm": 0.3517463803291321, | |
| "learning_rate": 5.629327647035843e-07, | |
| "loss": 0.0427, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.325842696629214, | |
| "grad_norm": 0.29264354705810547, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.0367, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.337078651685394, | |
| "grad_norm": 0.26809367537498474, | |
| "learning_rate": 5.27276800809875e-07, | |
| "loss": 0.0419, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.348314606741573, | |
| "grad_norm": 0.3042603135108948, | |
| "learning_rate": 5.098621211969224e-07, | |
| "loss": 0.0426, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.359550561797753, | |
| "grad_norm": 0.3065764904022217, | |
| "learning_rate": 4.92724414438771e-07, | |
| "loss": 0.0377, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.370786516853933, | |
| "grad_norm": 0.3297325074672699, | |
| "learning_rate": 4.758647376699033e-07, | |
| "loss": 0.0414, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.382022471910112, | |
| "grad_norm": 0.30828678607940674, | |
| "learning_rate": 4.5928413087459325e-07, | |
| "loss": 0.04, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.393258426966292, | |
| "grad_norm": 0.3171032965183258, | |
| "learning_rate": 4.4298361682277355e-07, | |
| "loss": 0.0396, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 4.404494382022472, | |
| "grad_norm": 0.2922931909561157, | |
| "learning_rate": 4.269642010069319e-07, | |
| "loss": 0.0373, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 4.415730337078652, | |
| "grad_norm": 0.30198633670806885, | |
| "learning_rate": 4.112268715800943e-07, | |
| "loss": 0.0454, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 4.426966292134831, | |
| "grad_norm": 0.2677709758281708, | |
| "learning_rate": 3.957725992948691e-07, | |
| "loss": 0.0378, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 4.438202247191011, | |
| "grad_norm": 0.3185364305973053, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.0484, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 4.449438202247191, | |
| "grad_norm": 0.33234983682632446, | |
| "learning_rate": 3.6571702179939604e-07, | |
| "loss": 0.0534, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 4.460674157303371, | |
| "grad_norm": 0.2978450059890747, | |
| "learning_rate": 3.511175705587433e-07, | |
| "loss": 0.041, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 4.47191011235955, | |
| "grad_norm": 0.30524784326553345, | |
| "learning_rate": 3.3680488428453005e-07, | |
| "loss": 0.0455, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 4.48314606741573, | |
| "grad_norm": 0.31283247470855713, | |
| "learning_rate": 3.227798458506637e-07, | |
| "loss": 0.0493, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 4.49438202247191, | |
| "grad_norm": 0.3062666356563568, | |
| "learning_rate": 3.0904332038757977e-07, | |
| "loss": 0.045, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 445, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 144825701957632.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |