{ "best_metric": 0.3522920608520508, "best_model_checkpoint": "miner_id_24/checkpoint-200", "epoch": 1.0540184453227932, "eval_steps": 50, "global_step": 200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005270092226613966, "grad_norm": 0.3474178910255432, "learning_rate": 1e-05, "loss": 0.971, "step": 1 }, { "epoch": 0.005270092226613966, "eval_loss": 2.119786024093628, "eval_runtime": 43.6721, "eval_samples_per_second": 7.327, "eval_steps_per_second": 1.832, "step": 1 }, { "epoch": 0.010540184453227932, "grad_norm": 0.3650890588760376, "learning_rate": 2e-05, "loss": 0.9456, "step": 2 }, { "epoch": 0.015810276679841896, "grad_norm": 0.3826043903827667, "learning_rate": 3e-05, "loss": 0.982, "step": 3 }, { "epoch": 0.021080368906455864, "grad_norm": 0.42453712224960327, "learning_rate": 4e-05, "loss": 0.9873, "step": 4 }, { "epoch": 0.026350461133069828, "grad_norm": 0.43494120240211487, "learning_rate": 5e-05, "loss": 0.985, "step": 5 }, { "epoch": 0.03162055335968379, "grad_norm": 0.41415026783943176, "learning_rate": 6e-05, "loss": 0.9258, "step": 6 }, { "epoch": 0.03689064558629776, "grad_norm": 0.3528198003768921, "learning_rate": 7e-05, "loss": 0.9, "step": 7 }, { "epoch": 0.04216073781291173, "grad_norm": 0.3321168124675751, "learning_rate": 8e-05, "loss": 0.8323, "step": 8 }, { "epoch": 0.04743083003952569, "grad_norm": 0.2890051603317261, "learning_rate": 9e-05, "loss": 0.7928, "step": 9 }, { "epoch": 0.052700922266139656, "grad_norm": 0.30663812160491943, "learning_rate": 0.0001, "loss": 0.706, "step": 10 }, { "epoch": 0.057971014492753624, "grad_norm": 0.3135335445404053, "learning_rate": 9.999316524962345e-05, "loss": 0.7286, "step": 11 }, { "epoch": 0.06324110671936758, "grad_norm": 0.32870882749557495, "learning_rate": 9.997266286704631e-05, "loss": 0.6647, "step": 12 }, { "epoch": 0.06851119894598155, "grad_norm": 0.30461087822914124, "learning_rate": 9.993849845741524e-05, "loss": 0.6717, "step": 13 }, { "epoch": 0.07378129117259552, "grad_norm": 0.28659287095069885, "learning_rate": 9.989068136093873e-05, "loss": 0.7105, "step": 14 }, { "epoch": 0.07905138339920949, "grad_norm": 0.2939622104167938, "learning_rate": 9.98292246503335e-05, "loss": 0.703, "step": 15 }, { "epoch": 0.08432147562582346, "grad_norm": 0.24742643535137177, "learning_rate": 9.975414512725057e-05, "loss": 0.6479, "step": 16 }, { "epoch": 0.08959156785243742, "grad_norm": 0.25503161549568176, "learning_rate": 9.966546331768191e-05, "loss": 0.6439, "step": 17 }, { "epoch": 0.09486166007905138, "grad_norm": 0.2335021197795868, "learning_rate": 9.956320346634876e-05, "loss": 0.5565, "step": 18 }, { "epoch": 0.10013175230566534, "grad_norm": 0.24304203689098358, "learning_rate": 9.944739353007344e-05, "loss": 0.6178, "step": 19 }, { "epoch": 0.10540184453227931, "grad_norm": 0.252155601978302, "learning_rate": 9.931806517013612e-05, "loss": 0.6081, "step": 20 }, { "epoch": 0.11067193675889328, "grad_norm": 0.21653133630752563, "learning_rate": 9.917525374361912e-05, "loss": 0.536, "step": 21 }, { "epoch": 0.11594202898550725, "grad_norm": 0.21302343904972076, "learning_rate": 9.901899829374047e-05, "loss": 0.4962, "step": 22 }, { "epoch": 0.12121212121212122, "grad_norm": 0.2472389191389084, "learning_rate": 9.884934153917997e-05, "loss": 0.5674, "step": 23 }, { "epoch": 0.12648221343873517, "grad_norm": 0.2353401929140091, "learning_rate": 9.86663298624003e-05, "loss": 0.5028, "step": 24 }, { "epoch": 0.13175230566534915, "grad_norm": 0.2188122570514679, "learning_rate": 9.847001329696653e-05, "loss": 0.479, "step": 25 }, { "epoch": 0.1370223978919631, "grad_norm": 0.23333917558193207, "learning_rate": 9.826044551386744e-05, "loss": 0.549, "step": 26 }, { "epoch": 0.1422924901185771, "grad_norm": 0.23388893902301788, "learning_rate": 9.803768380684242e-05, "loss": 0.5755, "step": 27 }, { "epoch": 0.14756258234519104, "grad_norm": 0.21468135714530945, "learning_rate": 9.780178907671789e-05, "loss": 0.4915, "step": 28 }, { "epoch": 0.152832674571805, "grad_norm": 0.2214045226573944, "learning_rate": 9.755282581475769e-05, "loss": 0.4243, "step": 29 }, { "epoch": 0.15810276679841898, "grad_norm": 0.19694066047668457, "learning_rate": 9.729086208503174e-05, "loss": 0.4503, "step": 30 }, { "epoch": 0.16337285902503293, "grad_norm": 0.21330076456069946, "learning_rate": 9.701596950580806e-05, "loss": 0.462, "step": 31 }, { "epoch": 0.1686429512516469, "grad_norm": 0.23391222953796387, "learning_rate": 9.672822322997305e-05, "loss": 0.4001, "step": 32 }, { "epoch": 0.17391304347826086, "grad_norm": 0.2338000237941742, "learning_rate": 9.642770192448536e-05, "loss": 0.377, "step": 33 }, { "epoch": 0.17918313570487485, "grad_norm": 0.21064437925815582, "learning_rate": 9.611448774886924e-05, "loss": 0.3712, "step": 34 }, { "epoch": 0.1844532279314888, "grad_norm": 0.2500576078891754, "learning_rate": 9.578866633275288e-05, "loss": 0.39, "step": 35 }, { "epoch": 0.18972332015810275, "grad_norm": 0.2840140163898468, "learning_rate": 9.545032675245813e-05, "loss": 0.349, "step": 36 }, { "epoch": 0.19499341238471674, "grad_norm": 4.862468719482422, "learning_rate": 9.509956150664796e-05, "loss": 1.6344, "step": 37 }, { "epoch": 0.2002635046113307, "grad_norm": 23.067012786865234, "learning_rate": 9.473646649103818e-05, "loss": 8.8915, "step": 38 }, { "epoch": 0.20553359683794467, "grad_norm": 11.134334564208984, "learning_rate": 9.43611409721806e-05, "loss": 5.7438, "step": 39 }, { "epoch": 0.21080368906455862, "grad_norm": 8.590675354003906, "learning_rate": 9.397368756032445e-05, "loss": 4.2111, "step": 40 }, { "epoch": 0.2160737812911726, "grad_norm": 6.200276851654053, "learning_rate": 9.357421218136386e-05, "loss": 3.0188, "step": 41 }, { "epoch": 0.22134387351778656, "grad_norm": 3.8253231048583984, "learning_rate": 9.316282404787871e-05, "loss": 2.1871, "step": 42 }, { "epoch": 0.22661396574440051, "grad_norm": 4.377999305725098, "learning_rate": 9.273963562927695e-05, "loss": 1.4974, "step": 43 }, { "epoch": 0.2318840579710145, "grad_norm": 2.9624979496002197, "learning_rate": 9.230476262104677e-05, "loss": 0.8439, "step": 44 }, { "epoch": 0.23715415019762845, "grad_norm": 2.5580661296844482, "learning_rate": 9.185832391312644e-05, "loss": 0.3837, "step": 45 }, { "epoch": 0.24242424242424243, "grad_norm": 1.1318681240081787, "learning_rate": 9.140044155740101e-05, "loss": 0.1051, "step": 46 }, { "epoch": 0.24769433465085638, "grad_norm": 5.065803527832031, "learning_rate": 9.093124073433463e-05, "loss": 0.1826, "step": 47 }, { "epoch": 0.25296442687747034, "grad_norm": 106.01213073730469, "learning_rate": 9.045084971874738e-05, "loss": 12.432, "step": 48 }, { "epoch": 0.25823451910408435, "grad_norm": 127.87001037597656, "learning_rate": 8.995939984474624e-05, "loss": 5.9008, "step": 49 }, { "epoch": 0.2635046113306983, "grad_norm": 18.64635467529297, "learning_rate": 8.945702546981969e-05, "loss": 1.5255, "step": 50 }, { "epoch": 0.2635046113306983, "eval_loss": 0.6848254203796387, "eval_runtime": 44.325, "eval_samples_per_second": 7.219, "eval_steps_per_second": 1.805, "step": 50 }, { "epoch": 0.26877470355731226, "grad_norm": 2.2442820072174072, "learning_rate": 8.894386393810563e-05, "loss": 0.8805, "step": 51 }, { "epoch": 0.2740447957839262, "grad_norm": 0.8949451446533203, "learning_rate": 8.842005554284296e-05, "loss": 0.725, "step": 52 }, { "epoch": 0.27931488801054016, "grad_norm": 0.5043036341667175, "learning_rate": 8.788574348801675e-05, "loss": 0.6606, "step": 53 }, { "epoch": 0.2845849802371542, "grad_norm": 0.34585607051849365, "learning_rate": 8.73410738492077e-05, "loss": 0.6511, "step": 54 }, { "epoch": 0.2898550724637681, "grad_norm": 0.2834174931049347, "learning_rate": 8.678619553365659e-05, "loss": 0.6144, "step": 55 }, { "epoch": 0.2951251646903821, "grad_norm": 0.29553502798080444, "learning_rate": 8.622126023955446e-05, "loss": 0.5787, "step": 56 }, { "epoch": 0.30039525691699603, "grad_norm": 0.2836796045303345, "learning_rate": 8.564642241456986e-05, "loss": 0.5745, "step": 57 }, { "epoch": 0.30566534914361, "grad_norm": 0.24634557962417603, "learning_rate": 8.506183921362443e-05, "loss": 0.5832, "step": 58 }, { "epoch": 0.310935441370224, "grad_norm": 0.246280238032341, "learning_rate": 8.44676704559283e-05, "loss": 0.5774, "step": 59 }, { "epoch": 0.31620553359683795, "grad_norm": 0.22724094986915588, "learning_rate": 8.386407858128706e-05, "loss": 0.5458, "step": 60 }, { "epoch": 0.3214756258234519, "grad_norm": 0.2311929315328598, "learning_rate": 8.32512286056924e-05, "loss": 0.6416, "step": 61 }, { "epoch": 0.32674571805006586, "grad_norm": 0.22324566543102264, "learning_rate": 8.262928807620843e-05, "loss": 0.552, "step": 62 }, { "epoch": 0.33201581027667987, "grad_norm": 0.20189306139945984, "learning_rate": 8.199842702516583e-05, "loss": 0.5467, "step": 63 }, { "epoch": 0.3372859025032938, "grad_norm": 0.20868399739265442, "learning_rate": 8.135881792367686e-05, "loss": 0.543, "step": 64 }, { "epoch": 0.3425559947299078, "grad_norm": 0.1898147463798523, "learning_rate": 8.07106356344834e-05, "loss": 0.5339, "step": 65 }, { "epoch": 0.34782608695652173, "grad_norm": 0.20499317348003387, "learning_rate": 8.005405736415126e-05, "loss": 0.4945, "step": 66 }, { "epoch": 0.3530961791831357, "grad_norm": 0.18965952098369598, "learning_rate": 7.938926261462366e-05, "loss": 0.5141, "step": 67 }, { "epoch": 0.3583662714097497, "grad_norm": 0.189968079328537, "learning_rate": 7.871643313414718e-05, "loss": 0.476, "step": 68 }, { "epoch": 0.36363636363636365, "grad_norm": 0.18742400407791138, "learning_rate": 7.803575286758364e-05, "loss": 0.5003, "step": 69 }, { "epoch": 0.3689064558629776, "grad_norm": 0.20146381855010986, "learning_rate": 7.734740790612136e-05, "loss": 0.5057, "step": 70 }, { "epoch": 0.37417654808959155, "grad_norm": 0.19327718019485474, "learning_rate": 7.66515864363997e-05, "loss": 0.4858, "step": 71 }, { "epoch": 0.3794466403162055, "grad_norm": 0.19030535221099854, "learning_rate": 7.594847868906076e-05, "loss": 0.4321, "step": 72 }, { "epoch": 0.3847167325428195, "grad_norm": 0.1938990354537964, "learning_rate": 7.52382768867422e-05, "loss": 0.4606, "step": 73 }, { "epoch": 0.38998682476943347, "grad_norm": 0.2044212967157364, "learning_rate": 7.452117519152542e-05, "loss": 0.4375, "step": 74 }, { "epoch": 0.3952569169960474, "grad_norm": 0.21626748144626617, "learning_rate": 7.379736965185368e-05, "loss": 0.4587, "step": 75 }, { "epoch": 0.4005270092226614, "grad_norm": 0.1907692849636078, "learning_rate": 7.30670581489344e-05, "loss": 0.3867, "step": 76 }, { "epoch": 0.4057971014492754, "grad_norm": 0.19831222295761108, "learning_rate": 7.233044034264034e-05, "loss": 0.4181, "step": 77 }, { "epoch": 0.41106719367588934, "grad_norm": 0.2016811966896057, "learning_rate": 7.158771761692464e-05, "loss": 0.3824, "step": 78 }, { "epoch": 0.4163372859025033, "grad_norm": 0.1924142986536026, "learning_rate": 7.083909302476453e-05, "loss": 0.3622, "step": 79 }, { "epoch": 0.42160737812911725, "grad_norm": 0.19537773728370667, "learning_rate": 7.008477123264848e-05, "loss": 0.4704, "step": 80 }, { "epoch": 0.4268774703557312, "grad_norm": 0.20724648237228394, "learning_rate": 6.932495846462261e-05, "loss": 0.3753, "step": 81 }, { "epoch": 0.4321475625823452, "grad_norm": 0.20100949704647064, "learning_rate": 6.855986244591104e-05, "loss": 0.3366, "step": 82 }, { "epoch": 0.43741765480895917, "grad_norm": 0.3353421688079834, "learning_rate": 6.778969234612584e-05, "loss": 0.3614, "step": 83 }, { "epoch": 0.4426877470355731, "grad_norm": 0.22832991182804108, "learning_rate": 6.701465872208216e-05, "loss": 0.3713, "step": 84 }, { "epoch": 0.4479578392621871, "grad_norm": 0.2740974724292755, "learning_rate": 6.623497346023418e-05, "loss": 0.289, "step": 85 }, { "epoch": 0.45322793148880103, "grad_norm": 5.220585823059082, "learning_rate": 6.545084971874738e-05, "loss": 1.1483, "step": 86 }, { "epoch": 0.45849802371541504, "grad_norm": 4.406704425811768, "learning_rate": 6.466250186922325e-05, "loss": 0.877, "step": 87 }, { "epoch": 0.463768115942029, "grad_norm": 3.400630474090576, "learning_rate": 6.387014543809223e-05, "loss": 0.4684, "step": 88 }, { "epoch": 0.46903820816864294, "grad_norm": 2.201713800430298, "learning_rate": 6.307399704769099e-05, "loss": 0.213, "step": 89 }, { "epoch": 0.4743083003952569, "grad_norm": 0.9386592507362366, "learning_rate": 6.227427435703997e-05, "loss": 0.0824, "step": 90 }, { "epoch": 0.4795783926218709, "grad_norm": 0.44810935854911804, "learning_rate": 6.147119600233758e-05, "loss": 0.0392, "step": 91 }, { "epoch": 0.48484848484848486, "grad_norm": 0.14609605073928833, "learning_rate": 6.066498153718735e-05, "loss": 0.013, "step": 92 }, { "epoch": 0.4901185770750988, "grad_norm": 0.10406555980443954, "learning_rate": 5.985585137257401e-05, "loss": 0.0066, "step": 93 }, { "epoch": 0.49538866930171277, "grad_norm": 0.15745611488819122, "learning_rate": 5.90440267166055e-05, "loss": 0.0054, "step": 94 }, { "epoch": 0.5006587615283268, "grad_norm": 2.1727049350738525, "learning_rate": 5.8229729514036705e-05, "loss": 0.7982, "step": 95 }, { "epoch": 0.5059288537549407, "grad_norm": 0.6171266436576843, "learning_rate": 5.74131823855921e-05, "loss": 0.683, "step": 96 }, { "epoch": 0.5111989459815547, "grad_norm": 0.3563093841075897, "learning_rate": 5.6594608567103456e-05, "loss": 0.5988, "step": 97 }, { "epoch": 0.5164690382081687, "grad_norm": 0.3043658435344696, "learning_rate": 5.577423184847932e-05, "loss": 0.5582, "step": 98 }, { "epoch": 0.5217391304347826, "grad_norm": 0.2405870407819748, "learning_rate": 5.495227651252315e-05, "loss": 0.6081, "step": 99 }, { "epoch": 0.5270092226613966, "grad_norm": 0.19620217382907867, "learning_rate": 5.4128967273616625e-05, "loss": 0.5297, "step": 100 }, { "epoch": 0.5270092226613966, "eval_loss": 0.40713223814964294, "eval_runtime": 44.3398, "eval_samples_per_second": 7.217, "eval_steps_per_second": 1.804, "step": 100 }, { "epoch": 0.5322793148880105, "grad_norm": 0.18272653222084045, "learning_rate": 5.330452921628497e-05, "loss": 0.5371, "step": 101 }, { "epoch": 0.5375494071146245, "grad_norm": 0.20360584557056427, "learning_rate": 5.247918773366112e-05, "loss": 0.5788, "step": 102 }, { "epoch": 0.5428194993412385, "grad_norm": 0.2235136181116104, "learning_rate": 5.165316846586541e-05, "loss": 0.5661, "step": 103 }, { "epoch": 0.5480895915678524, "grad_norm": 0.22581040859222412, "learning_rate": 5.0826697238317935e-05, "loss": 0.5207, "step": 104 }, { "epoch": 0.5533596837944664, "grad_norm": 0.20246632397174835, "learning_rate": 5e-05, "loss": 0.5348, "step": 105 }, { "epoch": 0.5586297760210803, "grad_norm": 0.19857186079025269, "learning_rate": 4.917330276168208e-05, "loss": 0.5835, "step": 106 }, { "epoch": 0.5638998682476943, "grad_norm": 0.18821997940540314, "learning_rate": 4.834683153413459e-05, "loss": 0.57, "step": 107 }, { "epoch": 0.5691699604743083, "grad_norm": 0.1922159343957901, "learning_rate": 4.7520812266338885e-05, "loss": 0.4962, "step": 108 }, { "epoch": 0.5744400527009222, "grad_norm": 0.21641184389591217, "learning_rate": 4.669547078371504e-05, "loss": 0.554, "step": 109 }, { "epoch": 0.5797101449275363, "grad_norm": 0.1995408684015274, "learning_rate": 4.5871032726383386e-05, "loss": 0.5369, "step": 110 }, { "epoch": 0.5849802371541502, "grad_norm": 0.20053839683532715, "learning_rate": 4.504772348747687e-05, "loss": 0.5335, "step": 111 }, { "epoch": 0.5902503293807642, "grad_norm": 0.19315959513187408, "learning_rate": 4.4225768151520694e-05, "loss": 0.4819, "step": 112 }, { "epoch": 0.5955204216073782, "grad_norm": 0.18426886200904846, "learning_rate": 4.3405391432896555e-05, "loss": 0.4814, "step": 113 }, { "epoch": 0.6007905138339921, "grad_norm": 0.20457793772220612, "learning_rate": 4.2586817614407895e-05, "loss": 0.5409, "step": 114 }, { "epoch": 0.6060606060606061, "grad_norm": 0.21226108074188232, "learning_rate": 4.17702704859633e-05, "loss": 0.5215, "step": 115 }, { "epoch": 0.61133069828722, "grad_norm": 0.2098385989665985, "learning_rate": 4.095597328339452e-05, "loss": 0.5202, "step": 116 }, { "epoch": 0.616600790513834, "grad_norm": 0.20233605802059174, "learning_rate": 4.0144148627425993e-05, "loss": 0.4843, "step": 117 }, { "epoch": 0.621870882740448, "grad_norm": 0.20992502570152283, "learning_rate": 3.933501846281267e-05, "loss": 0.5322, "step": 118 }, { "epoch": 0.6271409749670619, "grad_norm": 0.21740905940532684, "learning_rate": 3.852880399766243e-05, "loss": 0.4599, "step": 119 }, { "epoch": 0.6324110671936759, "grad_norm": 0.19706928730010986, "learning_rate": 3.772572564296005e-05, "loss": 0.4223, "step": 120 }, { "epoch": 0.6376811594202898, "grad_norm": 0.193538635969162, "learning_rate": 3.6926002952309016e-05, "loss": 0.4523, "step": 121 }, { "epoch": 0.6429512516469038, "grad_norm": 0.19901004433631897, "learning_rate": 3.612985456190778e-05, "loss": 0.4568, "step": 122 }, { "epoch": 0.6482213438735178, "grad_norm": 0.1916804164648056, "learning_rate": 3.533749813077677e-05, "loss": 0.4613, "step": 123 }, { "epoch": 0.6534914361001317, "grad_norm": 0.20516444742679596, "learning_rate": 3.4549150281252636e-05, "loss": 0.4528, "step": 124 }, { "epoch": 0.6587615283267457, "grad_norm": 0.20422868430614471, "learning_rate": 3.3765026539765834e-05, "loss": 0.4328, "step": 125 }, { "epoch": 0.6640316205533597, "grad_norm": 0.19037172198295593, "learning_rate": 3.298534127791785e-05, "loss": 0.3438, "step": 126 }, { "epoch": 0.6693017127799736, "grad_norm": 0.21251803636550903, "learning_rate": 3.221030765387417e-05, "loss": 0.3898, "step": 127 }, { "epoch": 0.6745718050065876, "grad_norm": 0.20027686655521393, "learning_rate": 3.144013755408895e-05, "loss": 0.3807, "step": 128 }, { "epoch": 0.6798418972332015, "grad_norm": 0.20919789373874664, "learning_rate": 3.0675041535377405e-05, "loss": 0.3387, "step": 129 }, { "epoch": 0.6851119894598156, "grad_norm": 0.203960120677948, "learning_rate": 2.991522876735154e-05, "loss": 0.2732, "step": 130 }, { "epoch": 0.6903820816864296, "grad_norm": 0.2608805298805237, "learning_rate": 2.916090697523549e-05, "loss": 0.2824, "step": 131 }, { "epoch": 0.6956521739130435, "grad_norm": 4.378519058227539, "learning_rate": 2.8412282383075363e-05, "loss": 0.4043, "step": 132 }, { "epoch": 0.7009222661396575, "grad_norm": 4.142157077789307, "learning_rate": 2.766955965735968e-05, "loss": 0.3387, "step": 133 }, { "epoch": 0.7061923583662714, "grad_norm": 2.9906086921691895, "learning_rate": 2.693294185106562e-05, "loss": 0.2003, "step": 134 }, { "epoch": 0.7114624505928854, "grad_norm": 1.246314525604248, "learning_rate": 2.6202630348146324e-05, "loss": 0.0853, "step": 135 }, { "epoch": 0.7167325428194994, "grad_norm": 0.6163206100463867, "learning_rate": 2.547882480847461e-05, "loss": 0.0441, "step": 136 }, { "epoch": 0.7220026350461133, "grad_norm": 0.3571086823940277, "learning_rate": 2.476172311325783e-05, "loss": 0.0195, "step": 137 }, { "epoch": 0.7272727272727273, "grad_norm": 0.17573851346969604, "learning_rate": 2.405152131093926e-05, "loss": 0.0114, "step": 138 }, { "epoch": 0.7325428194993412, "grad_norm": 0.5237570405006409, "learning_rate": 2.3348413563600325e-05, "loss": 0.0196, "step": 139 }, { "epoch": 0.7378129117259552, "grad_norm": 0.1536901593208313, "learning_rate": 2.2652592093878666e-05, "loss": 0.0084, "step": 140 }, { "epoch": 0.7430830039525692, "grad_norm": 0.11287213861942291, "learning_rate": 2.196424713241637e-05, "loss": 0.0025, "step": 141 }, { "epoch": 0.7483530961791831, "grad_norm": 0.24674616754055023, "learning_rate": 2.128356686585282e-05, "loss": 0.6079, "step": 142 }, { "epoch": 0.7536231884057971, "grad_norm": 0.22641393542289734, "learning_rate": 2.061073738537635e-05, "loss": 0.5214, "step": 143 }, { "epoch": 0.758893280632411, "grad_norm": 0.22398772835731506, "learning_rate": 1.9945942635848748e-05, "loss": 0.5228, "step": 144 }, { "epoch": 0.764163372859025, "grad_norm": 0.19416695833206177, "learning_rate": 1.928936436551661e-05, "loss": 0.5505, "step": 145 }, { "epoch": 0.769433465085639, "grad_norm": 0.207665354013443, "learning_rate": 1.8641182076323148e-05, "loss": 0.5079, "step": 146 }, { "epoch": 0.7747035573122529, "grad_norm": 0.18842846155166626, "learning_rate": 1.800157297483417e-05, "loss": 0.4972, "step": 147 }, { "epoch": 0.7799736495388669, "grad_norm": 0.19785141944885254, "learning_rate": 1.7370711923791567e-05, "loss": 0.5284, "step": 148 }, { "epoch": 0.7852437417654808, "grad_norm": 0.1773282289505005, "learning_rate": 1.6748771394307585e-05, "loss": 0.554, "step": 149 }, { "epoch": 0.7905138339920948, "grad_norm": 0.173191100358963, "learning_rate": 1.6135921418712956e-05, "loss": 0.4732, "step": 150 }, { "epoch": 0.7905138339920948, "eval_loss": 0.3699115514755249, "eval_runtime": 44.341, "eval_samples_per_second": 7.217, "eval_steps_per_second": 1.804, "step": 150 }, { "epoch": 0.7957839262187089, "grad_norm": 0.1772385835647583, "learning_rate": 1.553232954407171e-05, "loss": 0.5482, "step": 151 }, { "epoch": 0.8010540184453228, "grad_norm": 0.19023720920085907, "learning_rate": 1.4938160786375572e-05, "loss": 0.4997, "step": 152 }, { "epoch": 0.8063241106719368, "grad_norm": 0.1944209337234497, "learning_rate": 1.435357758543015e-05, "loss": 0.5175, "step": 153 }, { "epoch": 0.8115942028985508, "grad_norm": 0.19694966077804565, "learning_rate": 1.3778739760445552e-05, "loss": 0.4798, "step": 154 }, { "epoch": 0.8168642951251647, "grad_norm": 0.21851637959480286, "learning_rate": 1.3213804466343421e-05, "loss": 0.4925, "step": 155 }, { "epoch": 0.8221343873517787, "grad_norm": 0.20972071588039398, "learning_rate": 1.2658926150792322e-05, "loss": 0.5677, "step": 156 }, { "epoch": 0.8274044795783926, "grad_norm": 0.19723309576511383, "learning_rate": 1.2114256511983274e-05, "loss": 0.4702, "step": 157 }, { "epoch": 0.8326745718050066, "grad_norm": 0.19331055879592896, "learning_rate": 1.157994445715706e-05, "loss": 0.4891, "step": 158 }, { "epoch": 0.8379446640316206, "grad_norm": 0.19478961825370789, "learning_rate": 1.1056136061894384e-05, "loss": 0.493, "step": 159 }, { "epoch": 0.8432147562582345, "grad_norm": 0.20830045640468597, "learning_rate": 1.0542974530180327e-05, "loss": 0.5294, "step": 160 }, { "epoch": 0.8484848484848485, "grad_norm": 0.20098073780536652, "learning_rate": 1.0040600155253765e-05, "loss": 0.5498, "step": 161 }, { "epoch": 0.8537549407114624, "grad_norm": 0.18631495535373688, "learning_rate": 9.549150281252633e-06, "loss": 0.4691, "step": 162 }, { "epoch": 0.8590250329380764, "grad_norm": 0.18213127553462982, "learning_rate": 9.068759265665384e-06, "loss": 0.4664, "step": 163 }, { "epoch": 0.8642951251646904, "grad_norm": 0.18953700363636017, "learning_rate": 8.599558442598998e-06, "loss": 0.479, "step": 164 }, { "epoch": 0.8695652173913043, "grad_norm": 0.20108190178871155, "learning_rate": 8.141676086873572e-06, "loss": 0.5032, "step": 165 }, { "epoch": 0.8748353096179183, "grad_norm": 0.1972150206565857, "learning_rate": 7.695237378953223e-06, "loss": 0.4849, "step": 166 }, { "epoch": 0.8801054018445322, "grad_norm": 0.21045711636543274, "learning_rate": 7.260364370723044e-06, "loss": 0.3986, "step": 167 }, { "epoch": 0.8853754940711462, "grad_norm": 0.20088282227516174, "learning_rate": 6.837175952121306e-06, "loss": 0.4368, "step": 168 }, { "epoch": 0.8906455862977603, "grad_norm": 0.20243582129478455, "learning_rate": 6.425787818636131e-06, "loss": 0.4064, "step": 169 }, { "epoch": 0.8959156785243741, "grad_norm": 0.21716879308223724, "learning_rate": 6.026312439675552e-06, "loss": 0.3939, "step": 170 }, { "epoch": 0.9011857707509882, "grad_norm": 0.21941210329532623, "learning_rate": 5.6388590278194096e-06, "loss": 0.433, "step": 171 }, { "epoch": 0.9064558629776021, "grad_norm": 0.2347586303949356, "learning_rate": 5.263533508961827e-06, "loss": 0.4003, "step": 172 }, { "epoch": 0.9117259552042161, "grad_norm": 0.243482768535614, "learning_rate": 4.900438493352055e-06, "loss": 0.4134, "step": 173 }, { "epoch": 0.9169960474308301, "grad_norm": 0.23444387316703796, "learning_rate": 4.549673247541875e-06, "loss": 0.3914, "step": 174 }, { "epoch": 0.922266139657444, "grad_norm": 0.2640478014945984, "learning_rate": 4.2113336672471245e-06, "loss": 0.4066, "step": 175 }, { "epoch": 0.927536231884058, "grad_norm": 0.2862168848514557, "learning_rate": 3.885512251130763e-06, "loss": 0.3366, "step": 176 }, { "epoch": 0.932806324110672, "grad_norm": 0.30694127082824707, "learning_rate": 3.5722980755146517e-06, "loss": 0.3186, "step": 177 }, { "epoch": 0.9380764163372859, "grad_norm": 0.4981575310230255, "learning_rate": 3.271776770026963e-06, "loss": 0.2575, "step": 178 }, { "epoch": 0.9433465085638999, "grad_norm": 0.32179486751556396, "learning_rate": 2.9840304941919415e-06, "loss": 0.0177, "step": 179 }, { "epoch": 0.9486166007905138, "grad_norm": 0.03157076612114906, "learning_rate": 2.7091379149682685e-06, "loss": 0.0019, "step": 180 }, { "epoch": 0.9538866930171278, "grad_norm": 0.03689328953623772, "learning_rate": 2.4471741852423237e-06, "loss": 0.0022, "step": 181 }, { "epoch": 0.9591567852437418, "grad_norm": 0.0321175679564476, "learning_rate": 2.1982109232821178e-06, "loss": 0.0021, "step": 182 }, { "epoch": 0.9644268774703557, "grad_norm": 0.051594413816928864, "learning_rate": 1.962316193157593e-06, "loss": 0.0028, "step": 183 }, { "epoch": 0.9696969696969697, "grad_norm": 0.03357087820768356, "learning_rate": 1.7395544861325718e-06, "loss": 0.0021, "step": 184 }, { "epoch": 0.9749670619235836, "grad_norm": 0.037764906883239746, "learning_rate": 1.5299867030334814e-06, "loss": 0.0023, "step": 185 }, { "epoch": 0.9802371541501976, "grad_norm": 0.04509095847606659, "learning_rate": 1.333670137599713e-06, "loss": 0.0028, "step": 186 }, { "epoch": 0.9855072463768116, "grad_norm": 0.03999536857008934, "learning_rate": 1.1506584608200367e-06, "loss": 0.0022, "step": 187 }, { "epoch": 0.9907773386034255, "grad_norm": 0.005908744875341654, "learning_rate": 9.810017062595322e-07, "loss": 0.0003, "step": 188 }, { "epoch": 0.9960474308300395, "grad_norm": 0.16437992453575134, "learning_rate": 8.247462563808817e-07, "loss": 0.4958, "step": 189 }, { "epoch": 1.0013175230566536, "grad_norm": 0.29984813928604126, "learning_rate": 6.819348298638839e-07, "loss": 0.4583, "step": 190 }, { "epoch": 1.0065876152832676, "grad_norm": 0.15918108820915222, "learning_rate": 5.526064699265753e-07, "loss": 0.5372, "step": 191 }, { "epoch": 1.0118577075098814, "grad_norm": 0.1785496026277542, "learning_rate": 4.367965336512403e-07, "loss": 0.5255, "step": 192 }, { "epoch": 1.0171277997364954, "grad_norm": 0.15855196118354797, "learning_rate": 3.3453668231809286e-07, "loss": 0.5145, "step": 193 }, { "epoch": 1.0223978919631094, "grad_norm": 0.15700498223304749, "learning_rate": 2.458548727494292e-07, "loss": 0.5179, "step": 194 }, { "epoch": 1.0276679841897234, "grad_norm": 0.1596696972846985, "learning_rate": 1.7077534966650766e-07, "loss": 0.4777, "step": 195 }, { "epoch": 1.0329380764163374, "grad_norm": 0.1559162437915802, "learning_rate": 1.0931863906127327e-07, "loss": 0.485, "step": 196 }, { "epoch": 1.0382081686429512, "grad_norm": 0.17101773619651794, "learning_rate": 6.150154258476315e-08, "loss": 0.4604, "step": 197 }, { "epoch": 1.0434782608695652, "grad_norm": 0.1645197570323944, "learning_rate": 2.7337132953697554e-08, "loss": 0.4939, "step": 198 }, { "epoch": 1.0487483530961792, "grad_norm": 0.16165442764759064, "learning_rate": 6.834750376549792e-09, "loss": 0.5027, "step": 199 }, { "epoch": 1.0540184453227932, "grad_norm": 0.172588050365448, "learning_rate": 0.0, "loss": 0.5012, "step": 200 }, { "epoch": 1.0540184453227932, "eval_loss": 0.3522920608520508, "eval_runtime": 44.4094, "eval_samples_per_second": 7.206, "eval_steps_per_second": 1.801, "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": 5.255509203694387e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }