| { | |
| "best_metric": 2.124297857284546, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 1.3468013468013469, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006734006734006734, | |
| "grad_norm": 41.3544807434082, | |
| "learning_rate": 1e-05, | |
| "loss": 7.9248, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006734006734006734, | |
| "eval_loss": 3.2169711589813232, | |
| "eval_runtime": 2.8689, | |
| "eval_samples_per_second": 87.142, | |
| "eval_steps_per_second": 21.96, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013468013468013467, | |
| "grad_norm": 28.686861038208008, | |
| "learning_rate": 2e-05, | |
| "loss": 8.6452, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020202020202020204, | |
| "grad_norm": 27.811328887939453, | |
| "learning_rate": 3e-05, | |
| "loss": 8.6372, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.026936026936026935, | |
| "grad_norm": 31.2685489654541, | |
| "learning_rate": 4e-05, | |
| "loss": 9.4807, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03367003367003367, | |
| "grad_norm": 29.774410247802734, | |
| "learning_rate": 5e-05, | |
| "loss": 9.1981, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 32.26310729980469, | |
| "learning_rate": 6e-05, | |
| "loss": 9.1509, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04713804713804714, | |
| "grad_norm": 32.215721130371094, | |
| "learning_rate": 7e-05, | |
| "loss": 8.9049, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05387205387205387, | |
| "grad_norm": 39.98481750488281, | |
| "learning_rate": 8e-05, | |
| "loss": 9.0405, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 38.171199798583984, | |
| "learning_rate": 9e-05, | |
| "loss": 8.9062, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06734006734006734, | |
| "grad_norm": 41.6307487487793, | |
| "learning_rate": 0.0001, | |
| "loss": 8.9717, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 81.0295181274414, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 9.4454, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 55.38478088378906, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 9.6935, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08754208754208755, | |
| "grad_norm": 58.545833587646484, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 9.377, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09427609427609428, | |
| "grad_norm": 56.10289001464844, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.326, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10101010101010101, | |
| "grad_norm": 68.74766540527344, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 9.111, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10774410774410774, | |
| "grad_norm": 75.04796600341797, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.8751, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11447811447811448, | |
| "grad_norm": 97.14585876464844, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 9.1319, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 167.249755859375, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 9.463, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12794612794612795, | |
| "grad_norm": 83.27363586425781, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 9.3951, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13468013468013468, | |
| "grad_norm": 108.85181427001953, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 9.9799, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1414141414141414, | |
| "grad_norm": 115.3881607055664, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 9.9686, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 161.6963348388672, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 10.32, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15488215488215487, | |
| "grad_norm": 115.52421569824219, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 10.5488, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 126.47930145263672, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 10.08, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16835016835016836, | |
| "grad_norm": 169.35169982910156, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 11.3427, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1750841750841751, | |
| "grad_norm": 155.81097412109375, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 9.5669, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 143.9490966796875, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 11.2954, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18855218855218855, | |
| "grad_norm": 194.0313720703125, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 9.9716, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19528619528619529, | |
| "grad_norm": 191.34754943847656, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 10.9662, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 296.2303771972656, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 12.9639, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.20875420875420875, | |
| "grad_norm": 225.1488037109375, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 11.8622, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21548821548821548, | |
| "grad_norm": 300.85137939453125, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 10.5222, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 245.97015380859375, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 10.9324, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.22895622895622897, | |
| "grad_norm": 214.2495574951172, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 11.4371, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2356902356902357, | |
| "grad_norm": 155.1671905517578, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 10.6151, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 201.09104919433594, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 13.8874, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.24915824915824916, | |
| "grad_norm": 178.4291534423828, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 20.125, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2558922558922559, | |
| "grad_norm": 79.65760040283203, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 10.4598, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26262626262626265, | |
| "grad_norm": 78.33348083496094, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 9.8332, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.26936026936026936, | |
| "grad_norm": 64.6243667602539, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 10.621, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2760942760942761, | |
| "grad_norm": 61.71767807006836, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 9.495, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 68.86141204833984, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 9.409, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2895622895622896, | |
| "grad_norm": 44.217159271240234, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 8.914, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 40.44387435913086, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 8.5301, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 37.031822204589844, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 8.5377, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.30976430976430974, | |
| "grad_norm": 37.07893753051758, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 8.7213, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3164983164983165, | |
| "grad_norm": 35.096614837646484, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 8.255, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 33.91702651977539, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 8.4549, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.32996632996632996, | |
| "grad_norm": 35.421138763427734, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 8.232, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "grad_norm": 31.6516170501709, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 8.7055, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "eval_loss": 2.513667106628418, | |
| "eval_runtime": 2.8719, | |
| "eval_samples_per_second": 87.05, | |
| "eval_steps_per_second": 21.937, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3434343434343434, | |
| "grad_norm": 28.783037185668945, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 8.5083, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3501683501683502, | |
| "grad_norm": 40.46855163574219, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 8.9477, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3569023569023569, | |
| "grad_norm": 32.94813537597656, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 8.5355, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 41.55287170410156, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 7.8743, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 38.73907470703125, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 8.1458, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3771043771043771, | |
| "grad_norm": 44.616390228271484, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 8.2309, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3838383838383838, | |
| "grad_norm": 41.109676361083984, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 7.8411, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39057239057239057, | |
| "grad_norm": 37.3735237121582, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 7.7347, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.39730639730639733, | |
| "grad_norm": 52.334190368652344, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 9.2392, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 60.413455963134766, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 8.5869, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4107744107744108, | |
| "grad_norm": 60.67982864379883, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 8.8272, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4175084175084175, | |
| "grad_norm": 59.488704681396484, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 7.7681, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 72.16001892089844, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 8.2773, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43097643097643096, | |
| "grad_norm": 63.35856246948242, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 8.3866, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4377104377104377, | |
| "grad_norm": 58.37825012207031, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 8.092, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 66.32896423339844, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 9.5055, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4511784511784512, | |
| "grad_norm": 62.555328369140625, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.6575, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.45791245791245794, | |
| "grad_norm": 63.12342071533203, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 9.7882, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.46464646464646464, | |
| "grad_norm": 74.8153305053711, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 11.6926, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4713804713804714, | |
| "grad_norm": 78.0501937866211, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 12.0121, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4781144781144781, | |
| "grad_norm": 71.63357543945312, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 9.493, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 75.53740692138672, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 12.6566, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.49158249158249157, | |
| "grad_norm": 63.0469856262207, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.5392, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4983164983164983, | |
| "grad_norm": 107.28878784179688, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 14.1815, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5050505050505051, | |
| "grad_norm": 33.66437530517578, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 9.0827, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5117845117845118, | |
| "grad_norm": 39.31285858154297, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 9.3797, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 42.36201858520508, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 9.1241, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 40.273399353027344, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 8.866, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.531986531986532, | |
| "grad_norm": 38.195892333984375, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 9.1844, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5387205387205387, | |
| "grad_norm": 38.57444763183594, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 8.4501, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 35.04957962036133, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 8.7472, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5521885521885522, | |
| "grad_norm": 40.535308837890625, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 8.1346, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5589225589225589, | |
| "grad_norm": 50.49781036376953, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 8.5581, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 35.70309829711914, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 8.0395, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5723905723905723, | |
| "grad_norm": 39.775901794433594, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 7.6986, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5791245791245792, | |
| "grad_norm": 48.53861999511719, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 7.506, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5858585858585859, | |
| "grad_norm": 45.2303466796875, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 8.1955, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 39.96479034423828, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 8.6909, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5993265993265994, | |
| "grad_norm": 39.05266189575195, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 8.1678, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 62.654075622558594, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 8.1967, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6127946127946128, | |
| "grad_norm": 36.26161193847656, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 9.1161, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6195286195286195, | |
| "grad_norm": 41.742767333984375, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 8.8142, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6262626262626263, | |
| "grad_norm": 38.4517936706543, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 8.5365, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.632996632996633, | |
| "grad_norm": 60.15699005126953, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 7.5725, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6397306397306397, | |
| "grad_norm": 39.90503692626953, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 8.1235, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 42.461055755615234, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 8.0874, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6531986531986532, | |
| "grad_norm": 54.22385787963867, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 9.0772, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6599326599326599, | |
| "grad_norm": 53.31112289428711, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 8.3677, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 104.74864196777344, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 8.7863, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "grad_norm": 56.153167724609375, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 8.6025, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "eval_loss": 2.187391519546509, | |
| "eval_runtime": 2.8561, | |
| "eval_samples_per_second": 87.531, | |
| "eval_steps_per_second": 22.058, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6801346801346801, | |
| "grad_norm": 53.02522659301758, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 7.8997, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 67.17434692382812, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 7.3731, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6936026936026936, | |
| "grad_norm": 59.72895431518555, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 7.0059, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7003367003367004, | |
| "grad_norm": 53.37847137451172, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 8.9512, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7070707070707071, | |
| "grad_norm": 70.16453552246094, | |
| "learning_rate": 5e-05, | |
| "loss": 8.3674, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7138047138047138, | |
| "grad_norm": 100.9383316040039, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 9.4618, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7205387205387206, | |
| "grad_norm": 68.87419128417969, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 8.5877, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 65.74427032470703, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 11.0717, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.734006734006734, | |
| "grad_norm": 81.13140869140625, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 10.0225, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 79.43911743164062, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 11.1211, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7474747474747475, | |
| "grad_norm": 96.10649871826172, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 14.4545, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7542087542087542, | |
| "grad_norm": 42.719512939453125, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 7.3726, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7609427609427609, | |
| "grad_norm": 43.786014556884766, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 8.2409, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 85.84500885009766, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 8.5336, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7744107744107744, | |
| "grad_norm": 29.495527267456055, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 8.626, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7811447811447811, | |
| "grad_norm": 31.59947395324707, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 8.436, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 35.00060272216797, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 8.8905, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7946127946127947, | |
| "grad_norm": 35.44921875, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 8.8152, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8013468013468014, | |
| "grad_norm": 24.749345779418945, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 8.3101, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 26.719274520874023, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 8.0112, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 26.525959014892578, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 8.3156, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8215488215488216, | |
| "grad_norm": 26.646484375, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 7.5943, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8282828282828283, | |
| "grad_norm": 26.210987091064453, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 8.1803, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.835016835016835, | |
| "grad_norm": 27.579328536987305, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 7.5849, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8417508417508418, | |
| "grad_norm": 22.659543991088867, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 8.0374, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 24.334983825683594, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 8.5726, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8552188552188552, | |
| "grad_norm": 28.293781280517578, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 8.3815, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8619528619528619, | |
| "grad_norm": 25.025714874267578, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 8.0767, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8686868686868687, | |
| "grad_norm": 36.11143112182617, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 8.2696, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8754208754208754, | |
| "grad_norm": 31.006874084472656, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 8.452, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8821548821548821, | |
| "grad_norm": 31.031539916992188, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 8.1796, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 34.35560989379883, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 7.8535, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8956228956228957, | |
| "grad_norm": 42.85440444946289, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 8.1856, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9023569023569024, | |
| "grad_norm": 41.15959167480469, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 8.546, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 44.42498016357422, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 7.0382, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9158249158249159, | |
| "grad_norm": 44.232608795166016, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 8.3631, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9225589225589226, | |
| "grad_norm": 44.95029830932617, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 7.7547, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 54.041622161865234, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 6.9658, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.936026936026936, | |
| "grad_norm": 41.09820556640625, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 8.1192, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9427609427609428, | |
| "grad_norm": 48.26551818847656, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 9.2408, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9494949494949495, | |
| "grad_norm": 57.503726959228516, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 8.8095, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9562289562289562, | |
| "grad_norm": 47.783905029296875, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 8.405, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 67.84696197509766, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 8.6565, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 56.10593795776367, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 9.0028, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9764309764309764, | |
| "grad_norm": 67.27642822265625, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 8.055, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9831649831649831, | |
| "grad_norm": 77.75454711914062, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 8.8995, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.98989898989899, | |
| "grad_norm": 83.39681243896484, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 13.9811, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9966329966329966, | |
| "grad_norm": 76.89947509765625, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 14.4035, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0033670033670035, | |
| "grad_norm": 46.255950927734375, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 8.0268, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "grad_norm": 31.87166404724121, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 7.7841, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "eval_loss": 2.1300928592681885, | |
| "eval_runtime": 2.8825, | |
| "eval_samples_per_second": 86.732, | |
| "eval_steps_per_second": 21.856, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0168350168350169, | |
| "grad_norm": 26.951465606689453, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 8.9946, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0235690235690236, | |
| "grad_norm": 28.541532516479492, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 8.7658, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 30.127155303955078, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 8.567, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 27.5192928314209, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 8.1126, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0437710437710437, | |
| "grad_norm": 29.30331039428711, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 7.7907, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0505050505050506, | |
| "grad_norm": 23.36063003540039, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 8.1318, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0572390572390573, | |
| "grad_norm": 24.307653427124023, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 7.5939, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.063973063973064, | |
| "grad_norm": 32.096920013427734, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 7.852, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0707070707070707, | |
| "grad_norm": 34.85822296142578, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 7.4267, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0774410774410774, | |
| "grad_norm": 26.823911666870117, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 7.8495, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0841750841750841, | |
| "grad_norm": 26.102447509765625, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 8.1106, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 38.69279861450195, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 8.4846, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0976430976430978, | |
| "grad_norm": 25.4783878326416, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 7.8454, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.1043771043771045, | |
| "grad_norm": 47.460662841796875, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 8.0084, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 87.432373046875, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 7.5847, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.1178451178451179, | |
| "grad_norm": 30.893407821655273, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 7.4723, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.1245791245791246, | |
| "grad_norm": 34.20063781738281, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 7.3435, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1313131313131313, | |
| "grad_norm": 34.71181869506836, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 7.2139, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.138047138047138, | |
| "grad_norm": 29.156110763549805, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 6.9136, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.144781144781145, | |
| "grad_norm": 43.48200225830078, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 7.9598, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 38.563331604003906, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 8.8993, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1582491582491583, | |
| "grad_norm": 69.24166870117188, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 7.8323, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.164983164983165, | |
| "grad_norm": 52.72025680541992, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 7.6203, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1717171717171717, | |
| "grad_norm": 40.148155212402344, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 6.4577, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1784511784511784, | |
| "grad_norm": 56.67806625366211, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 6.1649, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 65.17906951904297, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 7.3353, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1919191919191918, | |
| "grad_norm": 52.39175033569336, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 6.4755, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1986531986531987, | |
| "grad_norm": 56.29236602783203, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 9.1806, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2053872053872055, | |
| "grad_norm": 54.4529914855957, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 9.822, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 60.61329650878906, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 7.9339, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2188552188552189, | |
| "grad_norm": 101.31444549560547, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 7.2201, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2255892255892256, | |
| "grad_norm": 62.9173583984375, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 7.7289, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2323232323232323, | |
| "grad_norm": 60.853878021240234, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 8.4224, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.239057239057239, | |
| "grad_norm": 65.28166198730469, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 10.1105, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2457912457912457, | |
| "grad_norm": 74.52928924560547, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.5693, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2525252525252526, | |
| "grad_norm": 77.29493713378906, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 10.1194, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2592592592592593, | |
| "grad_norm": 26.358734130859375, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 7.5299, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.265993265993266, | |
| "grad_norm": 30.440404891967773, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 6.9893, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 73.4139175415039, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 8.2673, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2794612794612794, | |
| "grad_norm": 35.271602630615234, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 7.7586, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2861952861952861, | |
| "grad_norm": 26.675586700439453, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 7.7927, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2929292929292928, | |
| "grad_norm": 29.052770614624023, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 8.3878, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2996632996632997, | |
| "grad_norm": 34.23065185546875, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 8.1041, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3063973063973064, | |
| "grad_norm": 24.850107192993164, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 7.4484, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3131313131313131, | |
| "grad_norm": 25.437028884887695, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 8.0154, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3198653198653199, | |
| "grad_norm": 43.03543472290039, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 7.3748, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3265993265993266, | |
| "grad_norm": 31.806800842285156, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 7.6694, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 38.405609130859375, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 7.5277, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.34006734006734, | |
| "grad_norm": 31.04942512512207, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 7.635, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "grad_norm": 32.92327117919922, | |
| "learning_rate": 0.0, | |
| "loss": 7.5463, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "eval_loss": 2.124297857284546, | |
| "eval_runtime": 2.8591, | |
| "eval_samples_per_second": 87.441, | |
| "eval_steps_per_second": 22.035, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4897778920521728e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |