| { | |
| "best_metric": 2.129817247390747, | |
| "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": 44.505859375, | |
| "learning_rate": 1e-05, | |
| "loss": 7.9248, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006734006734006734, | |
| "eval_loss": 3.2169711589813232, | |
| "eval_runtime": 2.8788, | |
| "eval_samples_per_second": 86.842, | |
| "eval_steps_per_second": 21.884, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013468013468013467, | |
| "grad_norm": 28.49445152282715, | |
| "learning_rate": 2e-05, | |
| "loss": 8.6452, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020202020202020204, | |
| "grad_norm": 28.70729637145996, | |
| "learning_rate": 3e-05, | |
| "loss": 8.6074, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.026936026936026935, | |
| "grad_norm": 32.5799446105957, | |
| "learning_rate": 4e-05, | |
| "loss": 9.4523, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03367003367003367, | |
| "grad_norm": 32.38400650024414, | |
| "learning_rate": 5e-05, | |
| "loss": 9.2279, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 33.52616500854492, | |
| "learning_rate": 6e-05, | |
| "loss": 9.1768, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04713804713804714, | |
| "grad_norm": 36.95421600341797, | |
| "learning_rate": 7e-05, | |
| "loss": 8.9234, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.05387205387205387, | |
| "grad_norm": 28.035762786865234, | |
| "learning_rate": 8e-05, | |
| "loss": 9.052, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 31.642139434814453, | |
| "learning_rate": 9e-05, | |
| "loss": 8.8748, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06734006734006734, | |
| "grad_norm": 30.315353393554688, | |
| "learning_rate": 0.0001, | |
| "loss": 9.0128, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 55.02390670776367, | |
| "learning_rate": 9.999316524962345e-05, | |
| "loss": 9.3137, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 44.553077697753906, | |
| "learning_rate": 9.997266286704631e-05, | |
| "loss": 9.6069, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08754208754208755, | |
| "grad_norm": 35.210914611816406, | |
| "learning_rate": 9.993849845741524e-05, | |
| "loss": 9.2278, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09427609427609428, | |
| "grad_norm": 40.97846221923828, | |
| "learning_rate": 9.989068136093873e-05, | |
| "loss": 9.1152, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10101010101010101, | |
| "grad_norm": 48.24727249145508, | |
| "learning_rate": 9.98292246503335e-05, | |
| "loss": 9.0056, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10774410774410774, | |
| "grad_norm": 52.307010650634766, | |
| "learning_rate": 9.975414512725057e-05, | |
| "loss": 8.6739, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11447811447811448, | |
| "grad_norm": 51.48130416870117, | |
| "learning_rate": 9.966546331768191e-05, | |
| "loss": 8.9487, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 53.33127212524414, | |
| "learning_rate": 9.956320346634876e-05, | |
| "loss": 9.124, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.12794612794612795, | |
| "grad_norm": 57.07469940185547, | |
| "learning_rate": 9.944739353007344e-05, | |
| "loss": 9.0992, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13468013468013468, | |
| "grad_norm": 54.28505325317383, | |
| "learning_rate": 9.931806517013612e-05, | |
| "loss": 9.6434, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1414141414141414, | |
| "grad_norm": 53.51996994018555, | |
| "learning_rate": 9.917525374361912e-05, | |
| "loss": 9.5509, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 63.32281494140625, | |
| "learning_rate": 9.901899829374047e-05, | |
| "loss": 9.8468, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15488215488215487, | |
| "grad_norm": 64.75834655761719, | |
| "learning_rate": 9.884934153917997e-05, | |
| "loss": 10.0868, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 66.29872131347656, | |
| "learning_rate": 9.86663298624003e-05, | |
| "loss": 9.5117, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.16835016835016836, | |
| "grad_norm": 85.2836685180664, | |
| "learning_rate": 9.847001329696653e-05, | |
| "loss": 10.7103, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1750841750841751, | |
| "grad_norm": 104.77082824707031, | |
| "learning_rate": 9.826044551386744e-05, | |
| "loss": 8.7021, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 113.63770294189453, | |
| "learning_rate": 9.803768380684242e-05, | |
| "loss": 10.5319, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.18855218855218855, | |
| "grad_norm": 77.97016906738281, | |
| "learning_rate": 9.780178907671789e-05, | |
| "loss": 8.8121, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19528619528619529, | |
| "grad_norm": 118.62873840332031, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 10.0043, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 104.96800231933594, | |
| "learning_rate": 9.729086208503174e-05, | |
| "loss": 11.5551, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.20875420875420875, | |
| "grad_norm": 91.46956634521484, | |
| "learning_rate": 9.701596950580806e-05, | |
| "loss": 9.9259, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21548821548821548, | |
| "grad_norm": 105.84339141845703, | |
| "learning_rate": 9.672822322997305e-05, | |
| "loss": 9.274, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 112.36212921142578, | |
| "learning_rate": 9.642770192448536e-05, | |
| "loss": 9.7014, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.22895622895622897, | |
| "grad_norm": 93.92436981201172, | |
| "learning_rate": 9.611448774886924e-05, | |
| "loss": 10.3099, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2356902356902357, | |
| "grad_norm": 85.32953643798828, | |
| "learning_rate": 9.578866633275288e-05, | |
| "loss": 9.6158, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 140.1669158935547, | |
| "learning_rate": 9.545032675245813e-05, | |
| "loss": 12.4101, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.24915824915824916, | |
| "grad_norm": 153.4528350830078, | |
| "learning_rate": 9.509956150664796e-05, | |
| "loss": 19.8508, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2558922558922559, | |
| "grad_norm": 66.96389770507812, | |
| "learning_rate": 9.473646649103818e-05, | |
| "loss": 9.7579, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26262626262626265, | |
| "grad_norm": 66.0352554321289, | |
| "learning_rate": 9.43611409721806e-05, | |
| "loss": 9.4209, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.26936026936026936, | |
| "grad_norm": 57.87456130981445, | |
| "learning_rate": 9.397368756032445e-05, | |
| "loss": 10.5969, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2760942760942761, | |
| "grad_norm": 53.16413879394531, | |
| "learning_rate": 9.357421218136386e-05, | |
| "loss": 9.3637, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 44.70361328125, | |
| "learning_rate": 9.316282404787871e-05, | |
| "loss": 9.338, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2895622895622896, | |
| "grad_norm": 46.938472747802734, | |
| "learning_rate": 9.273963562927695e-05, | |
| "loss": 8.9201, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 44.92363357543945, | |
| "learning_rate": 9.230476262104677e-05, | |
| "loss": 8.4643, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 54.77400588989258, | |
| "learning_rate": 9.185832391312644e-05, | |
| "loss": 8.543, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.30976430976430974, | |
| "grad_norm": 48.33302307128906, | |
| "learning_rate": 9.140044155740101e-05, | |
| "loss": 8.5842, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3164983164983165, | |
| "grad_norm": 41.1914176940918, | |
| "learning_rate": 9.093124073433463e-05, | |
| "loss": 8.3312, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 53.75209426879883, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 8.489, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.32996632996632996, | |
| "grad_norm": 56.443546295166016, | |
| "learning_rate": 8.995939984474624e-05, | |
| "loss": 8.3052, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "grad_norm": 61.102149963378906, | |
| "learning_rate": 8.945702546981969e-05, | |
| "loss": 8.6635, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "eval_loss": 2.5689687728881836, | |
| "eval_runtime": 2.8663, | |
| "eval_samples_per_second": 87.22, | |
| "eval_steps_per_second": 21.979, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3434343434343434, | |
| "grad_norm": 57.70844650268555, | |
| "learning_rate": 8.894386393810563e-05, | |
| "loss": 8.5058, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3501683501683502, | |
| "grad_norm": 54.68817901611328, | |
| "learning_rate": 8.842005554284296e-05, | |
| "loss": 9.2088, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3569023569023569, | |
| "grad_norm": 61.35806655883789, | |
| "learning_rate": 8.788574348801675e-05, | |
| "loss": 8.6844, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 51.0438232421875, | |
| "learning_rate": 8.73410738492077e-05, | |
| "loss": 8.1095, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 52.885528564453125, | |
| "learning_rate": 8.678619553365659e-05, | |
| "loss": 8.4086, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3771043771043771, | |
| "grad_norm": 54.405948638916016, | |
| "learning_rate": 8.622126023955446e-05, | |
| "loss": 8.3679, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3838383838383838, | |
| "grad_norm": 56.03476333618164, | |
| "learning_rate": 8.564642241456986e-05, | |
| "loss": 8.1032, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39057239057239057, | |
| "grad_norm": 67.06974029541016, | |
| "learning_rate": 8.506183921362443e-05, | |
| "loss": 8.0852, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.39730639730639733, | |
| "grad_norm": 80.96895599365234, | |
| "learning_rate": 8.44676704559283e-05, | |
| "loss": 9.476, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 104.01045227050781, | |
| "learning_rate": 8.386407858128706e-05, | |
| "loss": 8.9779, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4107744107744108, | |
| "grad_norm": 86.21642303466797, | |
| "learning_rate": 8.32512286056924e-05, | |
| "loss": 9.0375, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4175084175084175, | |
| "grad_norm": 124.29145050048828, | |
| "learning_rate": 8.262928807620843e-05, | |
| "loss": 7.9358, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 87.45722961425781, | |
| "learning_rate": 8.199842702516583e-05, | |
| "loss": 8.5587, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43097643097643096, | |
| "grad_norm": 82.48189544677734, | |
| "learning_rate": 8.135881792367686e-05, | |
| "loss": 8.2746, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.4377104377104377, | |
| "grad_norm": 59.54527282714844, | |
| "learning_rate": 8.07106356344834e-05, | |
| "loss": 8.1214, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 98.18574523925781, | |
| "learning_rate": 8.005405736415126e-05, | |
| "loss": 9.106, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.4511784511784512, | |
| "grad_norm": 96.0837173461914, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 10.37, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.45791245791245794, | |
| "grad_norm": 73.61067962646484, | |
| "learning_rate": 7.871643313414718e-05, | |
| "loss": 9.3572, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.46464646464646464, | |
| "grad_norm": 99.51297760009766, | |
| "learning_rate": 7.803575286758364e-05, | |
| "loss": 11.2465, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4713804713804714, | |
| "grad_norm": 102.87274169921875, | |
| "learning_rate": 7.734740790612136e-05, | |
| "loss": 11.7265, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4781144781144781, | |
| "grad_norm": 78.58678436279297, | |
| "learning_rate": 7.66515864363997e-05, | |
| "loss": 9.4691, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 86.38130950927734, | |
| "learning_rate": 7.594847868906076e-05, | |
| "loss": 12.6936, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.49158249158249157, | |
| "grad_norm": 65.78684997558594, | |
| "learning_rate": 7.52382768867422e-05, | |
| "loss": 10.6037, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4983164983164983, | |
| "grad_norm": 99.17400360107422, | |
| "learning_rate": 7.452117519152542e-05, | |
| "loss": 13.952, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5050505050505051, | |
| "grad_norm": 43.33572006225586, | |
| "learning_rate": 7.379736965185368e-05, | |
| "loss": 9.2769, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5117845117845118, | |
| "grad_norm": 46.7730827331543, | |
| "learning_rate": 7.30670581489344e-05, | |
| "loss": 9.4566, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 44.58665466308594, | |
| "learning_rate": 7.233044034264034e-05, | |
| "loss": 9.0467, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 41.195899963378906, | |
| "learning_rate": 7.158771761692464e-05, | |
| "loss": 8.9442, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.531986531986532, | |
| "grad_norm": 37.08003616333008, | |
| "learning_rate": 7.083909302476453e-05, | |
| "loss": 9.1218, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5387205387205387, | |
| "grad_norm": 34.10331344604492, | |
| "learning_rate": 7.008477123264848e-05, | |
| "loss": 8.5124, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 50.114322662353516, | |
| "learning_rate": 6.932495846462261e-05, | |
| "loss": 8.7382, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.5521885521885522, | |
| "grad_norm": 36.36222457885742, | |
| "learning_rate": 6.855986244591104e-05, | |
| "loss": 8.0493, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.5589225589225589, | |
| "grad_norm": 35.058753967285156, | |
| "learning_rate": 6.778969234612584e-05, | |
| "loss": 8.5914, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 30.800535202026367, | |
| "learning_rate": 6.701465872208216e-05, | |
| "loss": 8.0645, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5723905723905723, | |
| "grad_norm": 36.42448043823242, | |
| "learning_rate": 6.623497346023418e-05, | |
| "loss": 7.615, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5791245791245792, | |
| "grad_norm": 32.7035026550293, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 7.4631, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5858585858585859, | |
| "grad_norm": 38.15718460083008, | |
| "learning_rate": 6.466250186922325e-05, | |
| "loss": 8.1278, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 32.37458419799805, | |
| "learning_rate": 6.387014543809223e-05, | |
| "loss": 8.7058, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5993265993265994, | |
| "grad_norm": 147.5338592529297, | |
| "learning_rate": 6.307399704769099e-05, | |
| "loss": 8.0849, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 33.05725860595703, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 8.1298, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6127946127946128, | |
| "grad_norm": 28.77727508544922, | |
| "learning_rate": 6.147119600233758e-05, | |
| "loss": 9.1215, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6195286195286195, | |
| "grad_norm": 37.2505989074707, | |
| "learning_rate": 6.066498153718735e-05, | |
| "loss": 8.828, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6262626262626263, | |
| "grad_norm": 41.58840560913086, | |
| "learning_rate": 5.985585137257401e-05, | |
| "loss": 8.4415, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.632996632996633, | |
| "grad_norm": 36.12533950805664, | |
| "learning_rate": 5.90440267166055e-05, | |
| "loss": 7.503, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6397306397306397, | |
| "grad_norm": 31.363981246948242, | |
| "learning_rate": 5.8229729514036705e-05, | |
| "loss": 8.0746, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 42.88101577758789, | |
| "learning_rate": 5.74131823855921e-05, | |
| "loss": 8.0774, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6531986531986532, | |
| "grad_norm": 45.69608688354492, | |
| "learning_rate": 5.6594608567103456e-05, | |
| "loss": 9.2396, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6599326599326599, | |
| "grad_norm": 41.514305114746094, | |
| "learning_rate": 5.577423184847932e-05, | |
| "loss": 8.3588, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 62.62977981567383, | |
| "learning_rate": 5.495227651252315e-05, | |
| "loss": 8.4766, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "grad_norm": 61.53374099731445, | |
| "learning_rate": 5.4128967273616625e-05, | |
| "loss": 8.4616, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "eval_loss": 2.1881213188171387, | |
| "eval_runtime": 2.8658, | |
| "eval_samples_per_second": 87.235, | |
| "eval_steps_per_second": 21.983, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6801346801346801, | |
| "grad_norm": 66.96289825439453, | |
| "learning_rate": 5.330452921628497e-05, | |
| "loss": 8.1581, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 62.631980895996094, | |
| "learning_rate": 5.247918773366112e-05, | |
| "loss": 7.3518, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6936026936026936, | |
| "grad_norm": 53.04157638549805, | |
| "learning_rate": 5.165316846586541e-05, | |
| "loss": 7.1308, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7003367003367004, | |
| "grad_norm": 61.25350570678711, | |
| "learning_rate": 5.0826697238317935e-05, | |
| "loss": 9.0629, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7070707070707071, | |
| "grad_norm": 76.54383087158203, | |
| "learning_rate": 5e-05, | |
| "loss": 8.5302, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7138047138047138, | |
| "grad_norm": 78.238037109375, | |
| "learning_rate": 4.917330276168208e-05, | |
| "loss": 9.7869, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7205387205387206, | |
| "grad_norm": 81.0560302734375, | |
| "learning_rate": 4.834683153413459e-05, | |
| "loss": 8.5041, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 89.94346618652344, | |
| "learning_rate": 4.7520812266338885e-05, | |
| "loss": 11.4137, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.734006734006734, | |
| "grad_norm": 59.336788177490234, | |
| "learning_rate": 4.669547078371504e-05, | |
| "loss": 10.3382, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 88.44305419921875, | |
| "learning_rate": 4.5871032726383386e-05, | |
| "loss": 11.4242, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7474747474747475, | |
| "grad_norm": 116.59869384765625, | |
| "learning_rate": 4.504772348747687e-05, | |
| "loss": 14.3023, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7542087542087542, | |
| "grad_norm": 78.62010192871094, | |
| "learning_rate": 4.4225768151520694e-05, | |
| "loss": 7.5227, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7609427609427609, | |
| "grad_norm": 65.48169708251953, | |
| "learning_rate": 4.3405391432896555e-05, | |
| "loss": 8.3943, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 41.65544891357422, | |
| "learning_rate": 4.2586817614407895e-05, | |
| "loss": 8.5943, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7744107744107744, | |
| "grad_norm": 30.506385803222656, | |
| "learning_rate": 4.17702704859633e-05, | |
| "loss": 8.713, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7811447811447811, | |
| "grad_norm": 57.344844818115234, | |
| "learning_rate": 4.095597328339452e-05, | |
| "loss": 8.4521, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 36.802001953125, | |
| "learning_rate": 4.0144148627425993e-05, | |
| "loss": 8.9562, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.7946127946127947, | |
| "grad_norm": 47.20869827270508, | |
| "learning_rate": 3.933501846281267e-05, | |
| "loss": 8.8025, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8013468013468014, | |
| "grad_norm": 44.41746139526367, | |
| "learning_rate": 3.852880399766243e-05, | |
| "loss": 8.4029, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 28.782100677490234, | |
| "learning_rate": 3.772572564296005e-05, | |
| "loss": 8.1085, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 52.63249588012695, | |
| "learning_rate": 3.6926002952309016e-05, | |
| "loss": 8.4604, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8215488215488216, | |
| "grad_norm": 63.66533660888672, | |
| "learning_rate": 3.612985456190778e-05, | |
| "loss": 7.6301, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.8282828282828283, | |
| "grad_norm": 27.8382568359375, | |
| "learning_rate": 3.533749813077677e-05, | |
| "loss": 8.2819, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.835016835016835, | |
| "grad_norm": 39.87235641479492, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 7.7911, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8417508417508418, | |
| "grad_norm": 39.811702728271484, | |
| "learning_rate": 3.3765026539765834e-05, | |
| "loss": 8.1535, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 42.682090759277344, | |
| "learning_rate": 3.298534127791785e-05, | |
| "loss": 8.7152, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.8552188552188552, | |
| "grad_norm": 36.36883544921875, | |
| "learning_rate": 3.221030765387417e-05, | |
| "loss": 8.5839, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8619528619528619, | |
| "grad_norm": 29.702289581298828, | |
| "learning_rate": 3.144013755408895e-05, | |
| "loss": 8.1993, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8686868686868687, | |
| "grad_norm": 38.97806930541992, | |
| "learning_rate": 3.0675041535377405e-05, | |
| "loss": 8.3726, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8754208754208754, | |
| "grad_norm": 37.44873809814453, | |
| "learning_rate": 2.991522876735154e-05, | |
| "loss": 8.5547, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8821548821548821, | |
| "grad_norm": 38.29060363769531, | |
| "learning_rate": 2.916090697523549e-05, | |
| "loss": 8.1124, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 37.99305725097656, | |
| "learning_rate": 2.8412282383075363e-05, | |
| "loss": 7.8713, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.8956228956228957, | |
| "grad_norm": 78.97618103027344, | |
| "learning_rate": 2.766955965735968e-05, | |
| "loss": 7.9646, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9023569023569024, | |
| "grad_norm": 57.971187591552734, | |
| "learning_rate": 2.693294185106562e-05, | |
| "loss": 8.4712, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 70.23067474365234, | |
| "learning_rate": 2.6202630348146324e-05, | |
| "loss": 7.0797, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9158249158249159, | |
| "grad_norm": 67.08696746826172, | |
| "learning_rate": 2.547882480847461e-05, | |
| "loss": 8.4507, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9225589225589226, | |
| "grad_norm": 62.27402877807617, | |
| "learning_rate": 2.476172311325783e-05, | |
| "loss": 7.6961, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 58.5566291809082, | |
| "learning_rate": 2.405152131093926e-05, | |
| "loss": 6.9208, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.936026936026936, | |
| "grad_norm": 54.90800094604492, | |
| "learning_rate": 2.3348413563600325e-05, | |
| "loss": 7.9884, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9427609427609428, | |
| "grad_norm": 66.88945007324219, | |
| "learning_rate": 2.2652592093878666e-05, | |
| "loss": 9.0444, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9494949494949495, | |
| "grad_norm": 53.771629333496094, | |
| "learning_rate": 2.196424713241637e-05, | |
| "loss": 8.6803, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9562289562289562, | |
| "grad_norm": 72.6031494140625, | |
| "learning_rate": 2.128356686585282e-05, | |
| "loss": 8.3973, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 98.00312042236328, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 8.7639, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 74.7289047241211, | |
| "learning_rate": 1.9945942635848748e-05, | |
| "loss": 8.6959, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9764309764309764, | |
| "grad_norm": 58.80105972290039, | |
| "learning_rate": 1.928936436551661e-05, | |
| "loss": 8.1601, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9831649831649831, | |
| "grad_norm": 66.25013732910156, | |
| "learning_rate": 1.8641182076323148e-05, | |
| "loss": 9.0378, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.98989898989899, | |
| "grad_norm": 99.00608825683594, | |
| "learning_rate": 1.800157297483417e-05, | |
| "loss": 13.2388, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.9966329966329966, | |
| "grad_norm": 105.44629669189453, | |
| "learning_rate": 1.7370711923791567e-05, | |
| "loss": 13.857, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0033670033670035, | |
| "grad_norm": 110.58183288574219, | |
| "learning_rate": 1.6748771394307585e-05, | |
| "loss": 8.1917, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "grad_norm": 47.514259338378906, | |
| "learning_rate": 1.6135921418712956e-05, | |
| "loss": 7.8683, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0101010101010102, | |
| "eval_loss": 2.129817247390747, | |
| "eval_runtime": 2.8626, | |
| "eval_samples_per_second": 87.332, | |
| "eval_steps_per_second": 22.008, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0168350168350169, | |
| "grad_norm": 35.46605682373047, | |
| "learning_rate": 1.553232954407171e-05, | |
| "loss": 9.02, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0235690235690236, | |
| "grad_norm": 49.0546989440918, | |
| "learning_rate": 1.4938160786375572e-05, | |
| "loss": 8.8339, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0303030303030303, | |
| "grad_norm": 30.65929412841797, | |
| "learning_rate": 1.435357758543015e-05, | |
| "loss": 8.5788, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.037037037037037, | |
| "grad_norm": 44.25896072387695, | |
| "learning_rate": 1.3778739760445552e-05, | |
| "loss": 8.1816, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0437710437710437, | |
| "grad_norm": 99.33442687988281, | |
| "learning_rate": 1.3213804466343421e-05, | |
| "loss": 7.829, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0505050505050506, | |
| "grad_norm": 111.64092254638672, | |
| "learning_rate": 1.2658926150792322e-05, | |
| "loss": 8.1602, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0572390572390573, | |
| "grad_norm": 50.77913284301758, | |
| "learning_rate": 1.2114256511983274e-05, | |
| "loss": 7.6392, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.063973063973064, | |
| "grad_norm": 108.10916900634766, | |
| "learning_rate": 1.157994445715706e-05, | |
| "loss": 7.8437, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0707070707070707, | |
| "grad_norm": 45.10155487060547, | |
| "learning_rate": 1.1056136061894384e-05, | |
| "loss": 7.446, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0774410774410774, | |
| "grad_norm": 42.05150604248047, | |
| "learning_rate": 1.0542974530180327e-05, | |
| "loss": 7.7714, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0841750841750841, | |
| "grad_norm": 56.60260009765625, | |
| "learning_rate": 1.0040600155253765e-05, | |
| "loss": 8.1302, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 44.20862579345703, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 8.5648, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.0976430976430978, | |
| "grad_norm": 36.54599380493164, | |
| "learning_rate": 9.068759265665384e-06, | |
| "loss": 7.989, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.1043771043771045, | |
| "grad_norm": 81.98597717285156, | |
| "learning_rate": 8.599558442598998e-06, | |
| "loss": 8.0401, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 40.687744140625, | |
| "learning_rate": 8.141676086873572e-06, | |
| "loss": 7.553, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.1178451178451179, | |
| "grad_norm": 36.427791595458984, | |
| "learning_rate": 7.695237378953223e-06, | |
| "loss": 7.5782, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.1245791245791246, | |
| "grad_norm": 45.611934661865234, | |
| "learning_rate": 7.260364370723044e-06, | |
| "loss": 7.5116, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1313131313131313, | |
| "grad_norm": 47.503997802734375, | |
| "learning_rate": 6.837175952121306e-06, | |
| "loss": 7.1602, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.138047138047138, | |
| "grad_norm": 58.62235641479492, | |
| "learning_rate": 6.425787818636131e-06, | |
| "loss": 7.0491, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.144781144781145, | |
| "grad_norm": 52.05401611328125, | |
| "learning_rate": 6.026312439675552e-06, | |
| "loss": 8.0497, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.1515151515151516, | |
| "grad_norm": 47.182411193847656, | |
| "learning_rate": 5.6388590278194096e-06, | |
| "loss": 9.016, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1582491582491583, | |
| "grad_norm": 51.34718704223633, | |
| "learning_rate": 5.263533508961827e-06, | |
| "loss": 7.6707, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.164983164983165, | |
| "grad_norm": 64.08531188964844, | |
| "learning_rate": 4.900438493352055e-06, | |
| "loss": 7.7479, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1717171717171717, | |
| "grad_norm": 62.90645980834961, | |
| "learning_rate": 4.549673247541875e-06, | |
| "loss": 6.7577, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.1784511784511784, | |
| "grad_norm": 54.919891357421875, | |
| "learning_rate": 4.2113336672471245e-06, | |
| "loss": 6.3705, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 55.40079116821289, | |
| "learning_rate": 3.885512251130763e-06, | |
| "loss": 7.2107, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.1919191919191918, | |
| "grad_norm": 68.25981140136719, | |
| "learning_rate": 3.5722980755146517e-06, | |
| "loss": 6.3771, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.1986531986531987, | |
| "grad_norm": 65.10997772216797, | |
| "learning_rate": 3.271776770026963e-06, | |
| "loss": 9.1602, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2053872053872055, | |
| "grad_norm": 60.717926025390625, | |
| "learning_rate": 2.9840304941919415e-06, | |
| "loss": 9.6875, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2121212121212122, | |
| "grad_norm": 71.13054656982422, | |
| "learning_rate": 2.7091379149682685e-06, | |
| "loss": 7.9383, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2188552188552189, | |
| "grad_norm": 70.5322265625, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 7.0618, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2255892255892256, | |
| "grad_norm": 74.00837707519531, | |
| "learning_rate": 2.1982109232821178e-06, | |
| "loss": 7.48, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2323232323232323, | |
| "grad_norm": 87.33047485351562, | |
| "learning_rate": 1.962316193157593e-06, | |
| "loss": 8.2148, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.239057239057239, | |
| "grad_norm": 201.9131317138672, | |
| "learning_rate": 1.7395544861325718e-06, | |
| "loss": 10.4071, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2457912457912457, | |
| "grad_norm": 84.5881576538086, | |
| "learning_rate": 1.5299867030334814e-06, | |
| "loss": 10.4813, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2525252525252526, | |
| "grad_norm": 86.1329574584961, | |
| "learning_rate": 1.333670137599713e-06, | |
| "loss": 10.4515, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2592592592592593, | |
| "grad_norm": 36.676177978515625, | |
| "learning_rate": 1.1506584608200367e-06, | |
| "loss": 7.6659, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.265993265993266, | |
| "grad_norm": 34.234840393066406, | |
| "learning_rate": 9.810017062595322e-07, | |
| "loss": 7.1829, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 69.66173553466797, | |
| "learning_rate": 8.247462563808817e-07, | |
| "loss": 8.3657, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.2794612794612794, | |
| "grad_norm": 31.290864944458008, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 7.9053, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.2861952861952861, | |
| "grad_norm": 72.21870422363281, | |
| "learning_rate": 5.526064699265753e-07, | |
| "loss": 7.9443, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.2929292929292928, | |
| "grad_norm": 38.420162200927734, | |
| "learning_rate": 4.367965336512403e-07, | |
| "loss": 8.5182, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.2996632996632997, | |
| "grad_norm": 40.208003997802734, | |
| "learning_rate": 3.3453668231809286e-07, | |
| "loss": 8.3081, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3063973063973064, | |
| "grad_norm": 40.76144027709961, | |
| "learning_rate": 2.458548727494292e-07, | |
| "loss": 7.4706, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3131313131313131, | |
| "grad_norm": 61.534549713134766, | |
| "learning_rate": 1.7077534966650766e-07, | |
| "loss": 8.1495, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3198653198653199, | |
| "grad_norm": 50.20437240600586, | |
| "learning_rate": 1.0931863906127327e-07, | |
| "loss": 7.4514, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3265993265993266, | |
| "grad_norm": 44.8299560546875, | |
| "learning_rate": 6.150154258476315e-08, | |
| "loss": 7.7406, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 49.02394104003906, | |
| "learning_rate": 2.7337132953697554e-08, | |
| "loss": 7.6097, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.34006734006734, | |
| "grad_norm": 48.80705642700195, | |
| "learning_rate": 6.834750376549792e-09, | |
| "loss": 7.7851, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "grad_norm": 42.12969207763672, | |
| "learning_rate": 0.0, | |
| "loss": 7.6933, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3468013468013469, | |
| "eval_loss": 2.1342430114746094, | |
| "eval_runtime": 2.8623, | |
| "eval_samples_per_second": 87.342, | |
| "eval_steps_per_second": 22.01, | |
| "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 | |
| } | |