| { | |
| "best_metric": 2.191477060317993, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "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": 40.721534729003906, | |
| "learning_rate": 1e-05, | |
| "loss": 7.9248, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006734006734006734, | |
| "eval_loss": 3.2169711589813232, | |
| "eval_runtime": 2.8844, | |
| "eval_samples_per_second": 86.672, | |
| "eval_steps_per_second": 21.841, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013468013468013467, | |
| "grad_norm": 28.784969329833984, | |
| "learning_rate": 2e-05, | |
| "loss": 8.6452, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020202020202020204, | |
| "grad_norm": 27.089784622192383, | |
| "learning_rate": 3e-05, | |
| "loss": 8.6249, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.026936026936026935, | |
| "grad_norm": 28.68646240234375, | |
| "learning_rate": 4e-05, | |
| "loss": 9.4648, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03367003367003367, | |
| "grad_norm": 30.744503021240234, | |
| "learning_rate": 5e-05, | |
| "loss": 9.1962, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 33.686580657958984, | |
| "learning_rate": 6e-05, | |
| "loss": 9.1235, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04713804713804714, | |
| "grad_norm": 39.73603057861328, | |
| "learning_rate": 7e-05, | |
| "loss": 8.8869, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05387205387205387, | |
| "grad_norm": 30.490251541137695, | |
| "learning_rate": 8e-05, | |
| "loss": 9.062, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 31.70163917541504, | |
| "learning_rate": 9e-05, | |
| "loss": 8.851, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06734006734006734, | |
| "grad_norm": 33.06393051147461, | |
| "learning_rate": 0.0001, | |
| "loss": 8.9462, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 47.97706985473633, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 9.2913, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 37.36377716064453, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 9.6004, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08754208754208755, | |
| "grad_norm": 36.07746887207031, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 9.2947, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09427609427609428, | |
| "grad_norm": 40.56977844238281, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.1926, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10101010101010101, | |
| "grad_norm": 47.98197937011719, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 8.9654, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10774410774410774, | |
| "grad_norm": 54.175865173339844, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.6639, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11447811447811448, | |
| "grad_norm": 40.78084945678711, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 8.8979, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 101.21937561035156, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 9.2463, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12794612794612795, | |
| "grad_norm": 45.776824951171875, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 9.0019, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13468013468013468, | |
| "grad_norm": 39.45632553100586, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 9.5273, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1414141414141414, | |
| "grad_norm": 41.68073272705078, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 9.3874, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 56.13813781738281, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 9.7636, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15488215488215487, | |
| "grad_norm": 47.58816146850586, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 9.9965, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 84.44255065917969, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 9.4044, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16835016835016836, | |
| "grad_norm": 92.03646850585938, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 10.5061, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1750841750841751, | |
| "grad_norm": 91.179931640625, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 8.6226, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 88.350341796875, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 10.0611, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18855218855218855, | |
| "grad_norm": 77.5590591430664, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 8.6264, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19528619528619529, | |
| "grad_norm": 85.84996032714844, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 9.5288, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 121.55877685546875, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 11.2303, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.20875420875420875, | |
| "grad_norm": 122.58518981933594, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 9.4041, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21548821548821548, | |
| "grad_norm": 118.59822845458984, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 8.7499, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 118.8906478881836, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 9.4721, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.22895622895622897, | |
| "grad_norm": 151.17039489746094, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 10.0351, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2356902356902357, | |
| "grad_norm": 153.11143493652344, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 9.3901, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 186.2895965576172, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 12.4539, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.24915824915824916, | |
| "grad_norm": 272.4237365722656, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 19.0214, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2558922558922559, | |
| "grad_norm": 57.660282135009766, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 10.0976, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26262626262626265, | |
| "grad_norm": 61.9688606262207, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 9.6742, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.26936026936026936, | |
| "grad_norm": 48.77880859375, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 10.6098, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2760942760942761, | |
| "grad_norm": 59.31803894042969, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 9.4109, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 42.05841827392578, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 9.2186, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2895622895622896, | |
| "grad_norm": 46.42253112792969, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 8.8901, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 38.007572174072266, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 8.3063, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 38.153770446777344, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 8.4282, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.30976430976430974, | |
| "grad_norm": 59.13198471069336, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 8.6525, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3164983164983165, | |
| "grad_norm": 43.109901428222656, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 8.3236, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 52.856529235839844, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 8.3935, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.32996632996632996, | |
| "grad_norm": 65.39897155761719, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 8.3373, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "grad_norm": 58.22595977783203, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 8.729, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "eval_loss": 2.5757181644439697, | |
| "eval_runtime": 2.8741, | |
| "eval_samples_per_second": 86.983, | |
| "eval_steps_per_second": 21.92, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3434343434343434, | |
| "grad_norm": 56.74869918823242, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 8.5832, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3501683501683502, | |
| "grad_norm": 120.69475555419922, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 9.0831, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3569023569023569, | |
| "grad_norm": 90.3545913696289, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 8.7051, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 59.1437873840332, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 8.1265, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 85.22001647949219, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 8.4988, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3771043771043771, | |
| "grad_norm": 106.94491577148438, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 8.5023, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3838383838383838, | |
| "grad_norm": 101.17305755615234, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 8.1557, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39057239057239057, | |
| "grad_norm": 126.14461517333984, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 8.039, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.39730639730639733, | |
| "grad_norm": 152.2769317626953, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 9.834, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 119.77828216552734, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 8.9833, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4107744107744108, | |
| "grad_norm": 113.63468170166016, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 9.3487, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4175084175084175, | |
| "grad_norm": 142.9735107421875, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 8.3782, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 137.7916717529297, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 8.9443, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43097643097643096, | |
| "grad_norm": 231.43089294433594, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 9.1096, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4377104377104377, | |
| "grad_norm": 116.13423919677734, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 8.3049, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 145.97251892089844, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 9.7356, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4511784511784512, | |
| "grad_norm": 163.04188537597656, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.6661, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.45791245791245794, | |
| "grad_norm": 162.80068969726562, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 9.8382, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.46464646464646464, | |
| "grad_norm": 146.3478240966797, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 11.777, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4713804713804714, | |
| "grad_norm": 195.50843811035156, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 12.0246, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4781144781144781, | |
| "grad_norm": 328.9975280761719, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 9.7281, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 194.7329864501953, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 12.9932, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.49158249158249157, | |
| "grad_norm": 154.91452026367188, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.7869, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4983164983164983, | |
| "grad_norm": 157.42726135253906, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 14.3655, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5050505050505051, | |
| "grad_norm": 68.41136169433594, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 9.1286, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5117845117845118, | |
| "grad_norm": 94.79463958740234, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 9.4578, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 65.79988861083984, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 9.1846, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 642.8189697265625, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 8.9449, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.531986531986532, | |
| "grad_norm": 60.4979248046875, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 9.353, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5387205387205387, | |
| "grad_norm": 64.75019073486328, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 8.7256, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 69.59507751464844, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 8.9943, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5521885521885522, | |
| "grad_norm": 55.43488693237305, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 8.3704, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5589225589225589, | |
| "grad_norm": 66.44117736816406, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 8.7222, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 50.78106689453125, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 8.182, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5723905723905723, | |
| "grad_norm": 81.48322296142578, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 7.8204, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5791245791245792, | |
| "grad_norm": 94.91535186767578, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 7.6156, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5858585858585859, | |
| "grad_norm": 65.08795928955078, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 8.135, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 70.78136444091797, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 8.7126, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5993265993265994, | |
| "grad_norm": 83.88176727294922, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 8.183, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 68.87261199951172, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 8.1776, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6127946127946128, | |
| "grad_norm": 80.12345886230469, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 9.1202, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6195286195286195, | |
| "grad_norm": 100.9736328125, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 8.8236, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6262626262626263, | |
| "grad_norm": 76.5873031616211, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 8.6311, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.632996632996633, | |
| "grad_norm": 69.47529602050781, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 7.5208, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6397306397306397, | |
| "grad_norm": 97.56402587890625, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 8.3195, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 191.1015167236328, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 8.2413, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6531986531986532, | |
| "grad_norm": 147.45555114746094, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 9.2333, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6599326599326599, | |
| "grad_norm": 378.5626525878906, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 8.5199, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 136.10995483398438, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 8.8421, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "grad_norm": 111.9849853515625, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 8.6364, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "eval_loss": 2.275372266769409, | |
| "eval_runtime": 2.8712, | |
| "eval_samples_per_second": 87.072, | |
| "eval_steps_per_second": 21.942, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6801346801346801, | |
| "grad_norm": 104.74996185302734, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 8.2437, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 170.68528747558594, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 7.8121, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6936026936026936, | |
| "grad_norm": 128.83071899414062, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 7.4491, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7003367003367004, | |
| "grad_norm": 169.3953399658203, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 9.2013, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7070707070707071, | |
| "grad_norm": 152.36329650878906, | |
| "learning_rate": 5e-05, | |
| "loss": 9.2372, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7138047138047138, | |
| "grad_norm": 371.57012939453125, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 10.0807, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7205387205387206, | |
| "grad_norm": 330.9497375488281, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 9.2531, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 135.43008422851562, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 11.8979, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.734006734006734, | |
| "grad_norm": 212.6822052001953, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 11.1007, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 265.6589660644531, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 12.6706, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7474747474747475, | |
| "grad_norm": 485.5006408691406, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 15.7342, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7542087542087542, | |
| "grad_norm": 60.81523132324219, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 7.5859, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7609427609427609, | |
| "grad_norm": 67.88556671142578, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 8.3849, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 74.4134521484375, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 8.4995, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7744107744107744, | |
| "grad_norm": 64.49516296386719, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 8.6711, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7811447811447811, | |
| "grad_norm": 50.79487609863281, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 8.4351, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 60.97396469116211, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 8.8463, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7946127946127947, | |
| "grad_norm": 61.452674865722656, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 8.8858, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8013468013468014, | |
| "grad_norm": 72.31233978271484, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 8.3229, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 52.79329299926758, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 8.0779, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 45.338565826416016, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 8.3196, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8215488215488216, | |
| "grad_norm": 49.60345458984375, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 7.5808, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8282828282828283, | |
| "grad_norm": 41.162105560302734, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 8.2727, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.835016835016835, | |
| "grad_norm": 55.009700775146484, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 7.8265, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8417508417508418, | |
| "grad_norm": 53.04917526245117, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 8.2202, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 59.29380416870117, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 8.7888, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8552188552188552, | |
| "grad_norm": 105.81053161621094, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 8.6596, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8619528619528619, | |
| "grad_norm": 50.204811096191406, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 8.2665, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8686868686868687, | |
| "grad_norm": 138.47659301757812, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 8.4394, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8754208754208754, | |
| "grad_norm": 77.91687774658203, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 8.6879, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8821548821548821, | |
| "grad_norm": 73.20861053466797, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 8.2701, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 103.92080688476562, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 8.0998, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8956228956228957, | |
| "grad_norm": 73.09806060791016, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 8.2809, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9023569023569024, | |
| "grad_norm": 68.96542358398438, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 8.5987, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 109.05390930175781, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 7.231, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9158249158249159, | |
| "grad_norm": 102.29364013671875, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 8.5788, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9225589225589226, | |
| "grad_norm": 117.64923858642578, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 7.9096, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 82.27930450439453, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 7.2842, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.936026936026936, | |
| "grad_norm": 104.97429656982422, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 8.3299, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9427609427609428, | |
| "grad_norm": 174.6889190673828, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 9.6813, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9494949494949495, | |
| "grad_norm": 96.52239990234375, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 9.4073, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9562289562289562, | |
| "grad_norm": 124.64993286132812, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 8.747, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 135.41732788085938, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 8.8288, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 114.33064270019531, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 9.3506, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9764309764309764, | |
| "grad_norm": 169.06642150878906, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 8.7891, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9831649831649831, | |
| "grad_norm": 183.9605255126953, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 9.5495, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.98989898989899, | |
| "grad_norm": 193.75, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 13.829, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9966329966329966, | |
| "grad_norm": 245.53201293945312, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 14.8054, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0033670033670035, | |
| "grad_norm": 95.73715209960938, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 8.2234, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "grad_norm": 65.2020492553711, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 8.0546, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "eval_loss": 2.191477060317993, | |
| "eval_runtime": 2.8654, | |
| "eval_samples_per_second": 87.246, | |
| "eval_steps_per_second": 21.986, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0168350168350169, | |
| "grad_norm": 49.005245208740234, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 9.1243, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0235690235690236, | |
| "grad_norm": 340.1611633300781, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 8.827, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 46.67660140991211, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 8.6263, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 48.98415756225586, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 8.1752, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0437710437710437, | |
| "grad_norm": 58.78938293457031, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 7.9085, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0505050505050506, | |
| "grad_norm": 46.95909881591797, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 8.2227, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0572390572390573, | |
| "grad_norm": 52.07844924926758, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 7.654, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.063973063973064, | |
| "grad_norm": 63.95090866088867, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 7.9441, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0707070707070707, | |
| "grad_norm": 55.50286102294922, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 7.5894, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0774410774410774, | |
| "grad_norm": 53.47132110595703, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 7.9313, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0841750841750841, | |
| "grad_norm": 58.0089111328125, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 8.3368, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 52.48853302001953, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 8.7225, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0976430976430978, | |
| "grad_norm": 49.74880599975586, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 8.0528, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.1043771043771045, | |
| "grad_norm": 70.78367614746094, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 8.2333, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 83.07978820800781, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 7.6789, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.1178451178451179, | |
| "grad_norm": 46.41506576538086, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 7.8204, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.1245791245791246, | |
| "grad_norm": 74.65797424316406, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 7.6823, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1313131313131313, | |
| "grad_norm": 53.99269485473633, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 7.4254, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.138047138047138, | |
| "grad_norm": 82.40629577636719, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 7.2957, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.144781144781145, | |
| "grad_norm": 106.8160400390625, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 8.3195, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 143.20787048339844, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 9.3182, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1582491582491583, | |
| "grad_norm": 74.27720642089844, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 8.0101, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.164983164983165, | |
| "grad_norm": 104.70465087890625, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 7.8224, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1717171717171717, | |
| "grad_norm": 86.0539321899414, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 6.8349, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1784511784511784, | |
| "grad_norm": 85.72328186035156, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 6.5888, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 121.05365753173828, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 7.6368, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1919191919191918, | |
| "grad_norm": 285.8055114746094, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 7.0478, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1986531986531987, | |
| "grad_norm": 101.4668960571289, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 9.6618, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2053872053872055, | |
| "grad_norm": 169.57122802734375, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 10.1816, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 114.09046173095703, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 8.4842, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2188552188552189, | |
| "grad_norm": 99.68009185791016, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 7.6734, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2255892255892256, | |
| "grad_norm": 96.5626220703125, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 8.5017, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2323232323232323, | |
| "grad_norm": 104.6897964477539, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 9.0208, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.239057239057239, | |
| "grad_norm": 438.927001953125, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 10.9404, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2457912457912457, | |
| "grad_norm": 140.24510192871094, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 11.6866, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2525252525252526, | |
| "grad_norm": 161.13954162597656, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 11.262, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2592592592592593, | |
| "grad_norm": 35.379051208496094, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 7.6659, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.265993265993266, | |
| "grad_norm": 49.997467041015625, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 7.3364, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 61.83149337768555, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 8.4766, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2794612794612794, | |
| "grad_norm": 52.71331787109375, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 8.0765, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2861952861952861, | |
| "grad_norm": 51.639102935791016, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 8.0975, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2929292929292928, | |
| "grad_norm": 45.26035690307617, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 8.5436, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2996632996632997, | |
| "grad_norm": 70.1972427368164, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 8.3621, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3063973063973064, | |
| "grad_norm": 79.8095703125, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 7.5448, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3131313131313131, | |
| "grad_norm": 53.064212799072266, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 8.1653, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3198653198653199, | |
| "grad_norm": 42.038055419921875, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 7.4892, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3265993265993266, | |
| "grad_norm": 46.75093078613281, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 7.8642, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 107.9854736328125, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 7.6966, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.34006734006734, | |
| "grad_norm": 57.46672821044922, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 7.8818, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "grad_norm": 73.759033203125, | |
| "learning_rate": 0.0, | |
| "loss": 7.871, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "eval_loss": 2.203819513320923, | |
| "eval_runtime": 2.8644, | |
| "eval_samples_per_second": 87.279, | |
| "eval_steps_per_second": 21.994, | |
| "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": 1 | |
| } | |
| }, | |
| "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 | |
| } | |