| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0965971459934138, |
| "eval_steps": 50, |
| "global_step": 500, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0021953896816684962, |
| "grad_norm": 63.43226149633802, |
| "learning_rate": 0.0, |
| "loss": 0.9349, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.0043907793633369925, |
| "grad_norm": 36.34002028523472, |
| "learning_rate": 7.2992700729927e-09, |
| "loss": 0.84, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.006586169045005488, |
| "grad_norm": 35.076123786065985, |
| "learning_rate": 1.45985401459854e-08, |
| "loss": 0.872, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.008781558726673985, |
| "grad_norm": 35.02301842850996, |
| "learning_rate": 2.1897810218978102e-08, |
| "loss": 0.851, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.010976948408342482, |
| "grad_norm": 27.48208643895236, |
| "learning_rate": 2.91970802919708e-08, |
| "loss": 0.8809, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.013172338090010977, |
| "grad_norm": 45.477345508153874, |
| "learning_rate": 3.64963503649635e-08, |
| "loss": 0.9287, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.015367727771679473, |
| "grad_norm": 56.90118203461904, |
| "learning_rate": 4.3795620437956203e-08, |
| "loss": 0.9007, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.01756311745334797, |
| "grad_norm": 46.42558807941359, |
| "learning_rate": 5.10948905109489e-08, |
| "loss": 0.9807, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.019758507135016465, |
| "grad_norm": 42.603618137740504, |
| "learning_rate": 5.83941605839416e-08, |
| "loss": 0.9002, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.021953896816684963, |
| "grad_norm": 31.238590091540356, |
| "learning_rate": 6.569343065693431e-08, |
| "loss": 0.8321, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.024149286498353458, |
| "grad_norm": 74.05095323835384, |
| "learning_rate": 7.2992700729927e-08, |
| "loss": 0.7992, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.026344676180021953, |
| "grad_norm": 37.54325755059656, |
| "learning_rate": 8.029197080291971e-08, |
| "loss": 0.8845, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.02854006586169045, |
| "grad_norm": 42.06630970394279, |
| "learning_rate": 8.759124087591241e-08, |
| "loss": 0.9717, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.030735455543358946, |
| "grad_norm": 46.71572391565178, |
| "learning_rate": 9.48905109489051e-08, |
| "loss": 0.868, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.03293084522502744, |
| "grad_norm": 47.53527688382936, |
| "learning_rate": 1.021897810218978e-07, |
| "loss": 0.9339, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.03512623490669594, |
| "grad_norm": 30.17176017806663, |
| "learning_rate": 1.0948905109489052e-07, |
| "loss": 0.8985, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.03732162458836443, |
| "grad_norm": 53.77317643602209, |
| "learning_rate": 1.167883211678832e-07, |
| "loss": 0.9014, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.03951701427003293, |
| "grad_norm": 37.069232452020195, |
| "learning_rate": 1.240875912408759e-07, |
| "loss": 0.8917, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.04171240395170143, |
| "grad_norm": 57.37762629699942, |
| "learning_rate": 1.3138686131386862e-07, |
| "loss": 0.8932, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.043907793633369926, |
| "grad_norm": 35.926837420049885, |
| "learning_rate": 1.386861313868613e-07, |
| "loss": 0.903, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.04610318331503842, |
| "grad_norm": 32.248278685728465, |
| "learning_rate": 1.45985401459854e-07, |
| "loss": 0.8963, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.048298572996706916, |
| "grad_norm": 29.26799952212303, |
| "learning_rate": 1.532846715328467e-07, |
| "loss": 0.8644, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.050493962678375415, |
| "grad_norm": 30.785195210511173, |
| "learning_rate": 1.6058394160583942e-07, |
| "loss": 0.8437, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.052689352360043906, |
| "grad_norm": 33.71197663612937, |
| "learning_rate": 1.678832116788321e-07, |
| "loss": 0.8897, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.054884742041712405, |
| "grad_norm": 35.96552084220703, |
| "learning_rate": 1.7518248175182481e-07, |
| "loss": 0.8967, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.0570801317233809, |
| "grad_norm": 41.50138484689535, |
| "learning_rate": 1.824817518248175e-07, |
| "loss": 0.9074, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.059275521405049394, |
| "grad_norm": 38.797876016859426, |
| "learning_rate": 1.897810218978102e-07, |
| "loss": 0.9287, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.06147091108671789, |
| "grad_norm": 32.19034408841759, |
| "learning_rate": 1.9708029197080292e-07, |
| "loss": 0.8716, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.06366630076838639, |
| "grad_norm": 51.76045022284941, |
| "learning_rate": 2.043795620437956e-07, |
| "loss": 0.8812, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.06586169045005488, |
| "grad_norm": 31.359275794152296, |
| "learning_rate": 2.116788321167883e-07, |
| "loss": 0.8557, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.06805708013172337, |
| "grad_norm": 30.678686692428276, |
| "learning_rate": 2.1897810218978103e-07, |
| "loss": 0.8758, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.07025246981339188, |
| "grad_norm": 32.41276845704741, |
| "learning_rate": 2.2627737226277372e-07, |
| "loss": 0.8713, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.07244785949506037, |
| "grad_norm": 28.950024669686137, |
| "learning_rate": 2.335766423357664e-07, |
| "loss": 0.8627, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.07464324917672886, |
| "grad_norm": 18.417431019302867, |
| "learning_rate": 2.408759124087591e-07, |
| "loss": 0.7561, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.07683863885839737, |
| "grad_norm": 30.826596106550824, |
| "learning_rate": 2.481751824817518e-07, |
| "loss": 0.9044, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.07903402854006586, |
| "grad_norm": 27.86009502146877, |
| "learning_rate": 2.5547445255474454e-07, |
| "loss": 0.9158, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.08122941822173436, |
| "grad_norm": 33.10171663561771, |
| "learning_rate": 2.6277372262773725e-07, |
| "loss": 0.8973, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.08342480790340286, |
| "grad_norm": 33.54784642550598, |
| "learning_rate": 2.700729927007299e-07, |
| "loss": 0.8854, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.08562019758507135, |
| "grad_norm": 31.94225571174728, |
| "learning_rate": 2.773722627737226e-07, |
| "loss": 0.8634, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.08781558726673985, |
| "grad_norm": 18.667616381853037, |
| "learning_rate": 2.846715328467153e-07, |
| "loss": 0.8365, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.09001097694840834, |
| "grad_norm": 33.72334370651476, |
| "learning_rate": 2.91970802919708e-07, |
| "loss": 0.8826, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.09220636663007684, |
| "grad_norm": 26.43734180083069, |
| "learning_rate": 2.9927007299270075e-07, |
| "loss": 0.8895, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.09440175631174534, |
| "grad_norm": 31.963757242976843, |
| "learning_rate": 3.065693430656934e-07, |
| "loss": 0.8623, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.09659714599341383, |
| "grad_norm": 21.0251515948571, |
| "learning_rate": 3.138686131386861e-07, |
| "loss": 0.7966, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.09879253567508232, |
| "grad_norm": 31.038625878390096, |
| "learning_rate": 3.2116788321167883e-07, |
| "loss": 0.8834, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.10098792535675083, |
| "grad_norm": 20.27262840408739, |
| "learning_rate": 3.284671532846715e-07, |
| "loss": 0.833, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.10318331503841932, |
| "grad_norm": 27.9353455887378, |
| "learning_rate": 3.357664233576642e-07, |
| "loss": 0.8401, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.10537870472008781, |
| "grad_norm": 24.712611262961317, |
| "learning_rate": 3.4306569343065697e-07, |
| "loss": 0.759, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.10757409440175632, |
| "grad_norm": 19.616432703856, |
| "learning_rate": 3.5036496350364963e-07, |
| "loss": 0.7371, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.10976948408342481, |
| "grad_norm": 34.3573339583084, |
| "learning_rate": 3.5766423357664234e-07, |
| "loss": 0.8114, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.10976948408342481, |
| "eval_accuracy": 0.564, |
| "eval_loss": 0.6804020404815674, |
| "eval_runtime": 51.4851, |
| "eval_samples_per_second": 9.712, |
| "eval_steps_per_second": 1.224, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.1119648737650933, |
| "grad_norm": 52.44727313113604, |
| "learning_rate": 3.64963503649635e-07, |
| "loss": 0.8058, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.1141602634467618, |
| "grad_norm": 29.635805263155994, |
| "learning_rate": 3.722627737226277e-07, |
| "loss": 0.7798, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.1163556531284303, |
| "grad_norm": 16.05086854109441, |
| "learning_rate": 3.795620437956204e-07, |
| "loss": 0.7593, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.11855104281009879, |
| "grad_norm": 29.823000715758084, |
| "learning_rate": 3.8686131386861313e-07, |
| "loss": 0.7561, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.1207464324917673, |
| "grad_norm": 14.127126775727975, |
| "learning_rate": 3.9416058394160584e-07, |
| "loss": 0.7271, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.12294182217343579, |
| "grad_norm": 21.593202970179426, |
| "learning_rate": 4.0145985401459856e-07, |
| "loss": 0.7764, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.1251372118551043, |
| "grad_norm": 14.951837963554619, |
| "learning_rate": 4.087591240875912e-07, |
| "loss": 0.6718, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.12733260153677278, |
| "grad_norm": 15.378678774753151, |
| "learning_rate": 4.160583941605839e-07, |
| "loss": 0.7698, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.12952799121844127, |
| "grad_norm": 24.73055485970318, |
| "learning_rate": 4.233576642335766e-07, |
| "loss": 0.7673, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.13172338090010977, |
| "grad_norm": 17.57321911038973, |
| "learning_rate": 4.306569343065693e-07, |
| "loss": 0.7157, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.13391877058177826, |
| "grad_norm": 16.77467268690982, |
| "learning_rate": 4.3795620437956206e-07, |
| "loss": 0.7432, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.13611416026344675, |
| "grad_norm": 27.713531401351116, |
| "learning_rate": 4.452554744525547e-07, |
| "loss": 0.8336, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.13830954994511527, |
| "grad_norm": 18.28886043174935, |
| "learning_rate": 4.5255474452554743e-07, |
| "loss": 0.6868, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.14050493962678376, |
| "grad_norm": 17.56165424803153, |
| "learning_rate": 4.5985401459854014e-07, |
| "loss": 0.8134, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.14270032930845225, |
| "grad_norm": 16.76906700049957, |
| "learning_rate": 4.671532846715328e-07, |
| "loss": 0.7071, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.14489571899012074, |
| "grad_norm": 22.604350303541462, |
| "learning_rate": 4.744525547445255e-07, |
| "loss": 0.7717, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.14709110867178923, |
| "grad_norm": 15.57570075158043, |
| "learning_rate": 4.817518248175182e-07, |
| "loss": 0.6858, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.14928649835345773, |
| "grad_norm": 23.99766199274266, |
| "learning_rate": 4.89051094890511e-07, |
| "loss": 0.6925, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.15148188803512624, |
| "grad_norm": 14.622811969684838, |
| "learning_rate": 4.963503649635036e-07, |
| "loss": 0.6772, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.15367727771679474, |
| "grad_norm": 11.873629512549707, |
| "learning_rate": 5.036496350364964e-07, |
| "loss": 0.7201, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.15587266739846323, |
| "grad_norm": 14.971650059257875, |
| "learning_rate": 5.109489051094891e-07, |
| "loss": 0.7144, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.15806805708013172, |
| "grad_norm": 14.89771761097792, |
| "learning_rate": 5.182481751824817e-07, |
| "loss": 0.6823, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.1602634467618002, |
| "grad_norm": 26.145706119308308, |
| "learning_rate": 5.255474452554745e-07, |
| "loss": 0.6815, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.16245883644346873, |
| "grad_norm": 25.218475252582063, |
| "learning_rate": 5.328467153284672e-07, |
| "loss": 0.6683, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.16465422612513722, |
| "grad_norm": 24.795553870705557, |
| "learning_rate": 5.401459854014598e-07, |
| "loss": 0.6869, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.1668496158068057, |
| "grad_norm": 20.06359220796973, |
| "learning_rate": 5.474452554744526e-07, |
| "loss": 0.6776, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.1690450054884742, |
| "grad_norm": 14.605622927723077, |
| "learning_rate": 5.547445255474452e-07, |
| "loss": 0.6169, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.1712403951701427, |
| "grad_norm": 13.398369259423399, |
| "learning_rate": 5.620437956204379e-07, |
| "loss": 0.5911, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.1734357848518112, |
| "grad_norm": 22.20812822444505, |
| "learning_rate": 5.693430656934306e-07, |
| "loss": 0.6641, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.1756311745334797, |
| "grad_norm": 11.530488794545967, |
| "learning_rate": 5.766423357664233e-07, |
| "loss": 0.6546, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.1778265642151482, |
| "grad_norm": 19.45472907398047, |
| "learning_rate": 5.83941605839416e-07, |
| "loss": 0.6346, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.1800219538968167, |
| "grad_norm": 14.402272095344948, |
| "learning_rate": 5.912408759124087e-07, |
| "loss": 0.6157, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.18221734357848518, |
| "grad_norm": 20.486016309108074, |
| "learning_rate": 5.985401459854015e-07, |
| "loss": 0.6353, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.18441273326015367, |
| "grad_norm": 21.724039606610585, |
| "learning_rate": 6.058394160583942e-07, |
| "loss": 0.6669, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.18660812294182216, |
| "grad_norm": 23.550204927202188, |
| "learning_rate": 6.131386861313868e-07, |
| "loss": 0.6573, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.18880351262349068, |
| "grad_norm": 25.850348423694676, |
| "learning_rate": 6.204379562043796e-07, |
| "loss": 0.6096, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.19099890230515917, |
| "grad_norm": 15.668802168044579, |
| "learning_rate": 6.277372262773722e-07, |
| "loss": 0.6074, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.19319429198682767, |
| "grad_norm": 11.15207045876449, |
| "learning_rate": 6.350364963503649e-07, |
| "loss": 0.5861, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.19538968166849616, |
| "grad_norm": 30.989418951503534, |
| "learning_rate": 6.423357664233577e-07, |
| "loss": 0.5803, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.19758507135016465, |
| "grad_norm": 37.87369273674589, |
| "learning_rate": 6.496350364963503e-07, |
| "loss": 0.5607, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.19978046103183314, |
| "grad_norm": 16.304752909448357, |
| "learning_rate": 6.56934306569343e-07, |
| "loss": 0.4855, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.20197585071350166, |
| "grad_norm": 46.813965107422554, |
| "learning_rate": 6.642335766423358e-07, |
| "loss": 0.5277, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.20417124039517015, |
| "grad_norm": 16.25834561476597, |
| "learning_rate": 6.715328467153284e-07, |
| "loss": 0.5671, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.20636663007683864, |
| "grad_norm": 24.571054002712096, |
| "learning_rate": 6.788321167883211e-07, |
| "loss": 0.5485, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.20856201975850713, |
| "grad_norm": 17.16346868917258, |
| "learning_rate": 6.861313868613139e-07, |
| "loss": 0.6168, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.21075740944017562, |
| "grad_norm": 24.939402022765716, |
| "learning_rate": 6.934306569343066e-07, |
| "loss": 0.5097, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.21295279912184412, |
| "grad_norm": 35.04290652380563, |
| "learning_rate": 7.007299270072993e-07, |
| "loss": 0.511, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.21514818880351264, |
| "grad_norm": 19.369038528744387, |
| "learning_rate": 7.080291970802919e-07, |
| "loss": 0.501, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.21734357848518113, |
| "grad_norm": 14.504342290526917, |
| "learning_rate": 7.153284671532847e-07, |
| "loss": 0.5576, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.21953896816684962, |
| "grad_norm": 22.831165380560154, |
| "learning_rate": 7.226277372262773e-07, |
| "loss": 0.6357, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.21953896816684962, |
| "eval_accuracy": 0.716, |
| "eval_loss": 0.5146071910858154, |
| "eval_runtime": 51.4544, |
| "eval_samples_per_second": 9.717, |
| "eval_steps_per_second": 1.224, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.2217343578485181, |
| "grad_norm": 17.214759546075644, |
| "learning_rate": 7.2992700729927e-07, |
| "loss": 0.5879, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.2239297475301866, |
| "grad_norm": 28.14453156116428, |
| "learning_rate": 7.372262773722628e-07, |
| "loss": 0.6077, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.2261251372118551, |
| "grad_norm": 36.81857394774189, |
| "learning_rate": 7.445255474452554e-07, |
| "loss": 0.5656, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.2283205268935236, |
| "grad_norm": 19.720256889573427, |
| "learning_rate": 7.518248175182481e-07, |
| "loss": 0.5779, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.2305159165751921, |
| "grad_norm": 13.404265085713362, |
| "learning_rate": 7.591240875912408e-07, |
| "loss": 0.6194, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.2327113062568606, |
| "grad_norm": 10.223644936756514, |
| "learning_rate": 7.664233576642335e-07, |
| "loss": 0.5653, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.2349066959385291, |
| "grad_norm": 11.802424231339492, |
| "learning_rate": 7.737226277372263e-07, |
| "loss": 0.5243, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.23710208562019758, |
| "grad_norm": 15.21599780255988, |
| "learning_rate": 7.81021897810219e-07, |
| "loss": 0.5237, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.23929747530186607, |
| "grad_norm": 9.817905688792953, |
| "learning_rate": 7.883211678832117e-07, |
| "loss": 0.5559, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.2414928649835346, |
| "grad_norm": 14.019441837913432, |
| "learning_rate": 7.956204379562043e-07, |
| "loss": 0.5634, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.24368825466520308, |
| "grad_norm": 11.287304916292086, |
| "learning_rate": 8.029197080291971e-07, |
| "loss": 0.5169, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.24588364434687157, |
| "grad_norm": 13.731826260486923, |
| "learning_rate": 8.102189781021898e-07, |
| "loss": 0.4922, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.24807903402854006, |
| "grad_norm": 10.127768690583599, |
| "learning_rate": 8.175182481751824e-07, |
| "loss": 0.4939, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.2502744237102086, |
| "grad_norm": 10.643603728793044, |
| "learning_rate": 8.248175182481751e-07, |
| "loss": 0.5433, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.2524698133918771, |
| "grad_norm": 16.423886102329853, |
| "learning_rate": 8.321167883211679e-07, |
| "loss": 0.4931, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.25466520307354557, |
| "grad_norm": 11.683580419342695, |
| "learning_rate": 8.394160583941605e-07, |
| "loss": 0.6014, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.25686059275521406, |
| "grad_norm": 9.964987145456663, |
| "learning_rate": 8.467153284671532e-07, |
| "loss": 0.5111, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.25905598243688255, |
| "grad_norm": 11.994223750763792, |
| "learning_rate": 8.540145985401459e-07, |
| "loss": 0.5427, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.26125137211855104, |
| "grad_norm": 17.08922954256959, |
| "learning_rate": 8.613138686131386e-07, |
| "loss": 0.5454, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.26344676180021953, |
| "grad_norm": 9.233224929973781, |
| "learning_rate": 8.686131386861314e-07, |
| "loss": 0.4425, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.265642151481888, |
| "grad_norm": 11.158793551879755, |
| "learning_rate": 8.759124087591241e-07, |
| "loss": 0.4592, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.2678375411635565, |
| "grad_norm": 9.984032512868838, |
| "learning_rate": 8.832116788321168e-07, |
| "loss": 0.4619, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.270032930845225, |
| "grad_norm": 13.6121955869425, |
| "learning_rate": 8.905109489051094e-07, |
| "loss": 0.5102, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.2722283205268935, |
| "grad_norm": 16.38565463082547, |
| "learning_rate": 8.978102189781022e-07, |
| "loss": 0.5414, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.27442371020856204, |
| "grad_norm": 13.72322184659144, |
| "learning_rate": 9.051094890510949e-07, |
| "loss": 0.4404, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.27661909989023054, |
| "grad_norm": 9.728231053042531, |
| "learning_rate": 9.124087591240875e-07, |
| "loss": 0.4891, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.278814489571899, |
| "grad_norm": 19.653134550591925, |
| "learning_rate": 9.197080291970803e-07, |
| "loss": 0.4911, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.2810098792535675, |
| "grad_norm": 15.233058767794994, |
| "learning_rate": 9.270072992700729e-07, |
| "loss": 0.4482, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.283205268935236, |
| "grad_norm": 14.85049447902684, |
| "learning_rate": 9.343065693430656e-07, |
| "loss": 0.4381, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.2854006586169045, |
| "grad_norm": 18.23588962213726, |
| "learning_rate": 9.416058394160583e-07, |
| "loss": 0.5267, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.287596048298573, |
| "grad_norm": 13.687385484203132, |
| "learning_rate": 9.48905109489051e-07, |
| "loss": 0.533, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.2897914379802415, |
| "grad_norm": 16.31348614065883, |
| "learning_rate": 9.562043795620438e-07, |
| "loss": 0.4592, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.29198682766191, |
| "grad_norm": 10.619144950897025, |
| "learning_rate": 9.635036496350364e-07, |
| "loss": 0.5337, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.29418221734357847, |
| "grad_norm": 10.964180780818202, |
| "learning_rate": 9.708029197080291e-07, |
| "loss": 0.531, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.29637760702524696, |
| "grad_norm": 9.68554990094299, |
| "learning_rate": 9.78102189781022e-07, |
| "loss": 0.4466, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.29857299670691545, |
| "grad_norm": 11.193131252386175, |
| "learning_rate": 9.854014598540146e-07, |
| "loss": 0.4518, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.300768386388584, |
| "grad_norm": 7.652184708715949, |
| "learning_rate": 9.927007299270073e-07, |
| "loss": 0.4514, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.3029637760702525, |
| "grad_norm": 9.282129557780532, |
| "learning_rate": 1e-06, |
| "loss": 0.4952, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.305159165751921, |
| "grad_norm": 9.514439267127576, |
| "learning_rate": 9.999983717412808e-07, |
| "loss": 0.5277, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.30735455543358947, |
| "grad_norm": 8.04642891691016, |
| "learning_rate": 9.999934869757278e-07, |
| "loss": 0.5323, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.30954994511525796, |
| "grad_norm": 14.24768033484951, |
| "learning_rate": 9.999853457351558e-07, |
| "loss": 0.4996, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.31174533479692645, |
| "grad_norm": 11.061860342712375, |
| "learning_rate": 9.999739480725893e-07, |
| "loss": 0.4849, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.31394072447859495, |
| "grad_norm": 15.443833167332631, |
| "learning_rate": 9.999592940622613e-07, |
| "loss": 0.4957, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.31613611416026344, |
| "grad_norm": 9.488802082671176, |
| "learning_rate": 9.999413837996137e-07, |
| "loss": 0.495, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.31833150384193193, |
| "grad_norm": 7.092185690265562, |
| "learning_rate": 9.999202174012972e-07, |
| "loss": 0.4189, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.3205268935236004, |
| "grad_norm": 8.422031666334862, |
| "learning_rate": 9.99895795005169e-07, |
| "loss": 0.4548, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.3227222832052689, |
| "grad_norm": 9.442638766406834, |
| "learning_rate": 9.99868116770293e-07, |
| "loss": 0.4385, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.32491767288693746, |
| "grad_norm": 7.687755924564992, |
| "learning_rate": 9.998371828769384e-07, |
| "loss": 0.4726, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.32711306256860595, |
| "grad_norm": 11.01823455774795, |
| "learning_rate": 9.99802993526579e-07, |
| "loss": 0.5062, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.32930845225027444, |
| "grad_norm": 8.828666504433839, |
| "learning_rate": 9.997655489418912e-07, |
| "loss": 0.5211, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.32930845225027444, |
| "eval_accuracy": 0.774, |
| "eval_loss": 0.4343988001346588, |
| "eval_runtime": 52.0075, |
| "eval_samples_per_second": 9.614, |
| "eval_steps_per_second": 1.211, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.33150384193194293, |
| "grad_norm": 9.826049631904308, |
| "learning_rate": 9.997248493667527e-07, |
| "loss": 0.5144, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.3336992316136114, |
| "grad_norm": 9.822644682623359, |
| "learning_rate": 9.996808950662413e-07, |
| "loss": 0.46, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.3358946212952799, |
| "grad_norm": 7.448251385258531, |
| "learning_rate": 9.99633686326633e-07, |
| "loss": 0.4837, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.3380900109769484, |
| "grad_norm": 8.66825069576008, |
| "learning_rate": 9.995832234554e-07, |
| "loss": 0.4442, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.3402854006586169, |
| "grad_norm": 9.445663098800049, |
| "learning_rate": 9.995295067812083e-07, |
| "loss": 0.4984, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.3424807903402854, |
| "grad_norm": 7.886446457789178, |
| "learning_rate": 9.99472536653917e-07, |
| "loss": 0.4888, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.3446761800219539, |
| "grad_norm": 6.5270562387929285, |
| "learning_rate": 9.994123134445746e-07, |
| "loss": 0.4608, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.3468715697036224, |
| "grad_norm": 8.44851437549508, |
| "learning_rate": 9.993488375454165e-07, |
| "loss": 0.4319, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.34906695938529086, |
| "grad_norm": 8.945522697664286, |
| "learning_rate": 9.992821093698636e-07, |
| "loss": 0.52, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.3512623490669594, |
| "grad_norm": 6.9004623022662965, |
| "learning_rate": 9.992121293525188e-07, |
| "loss": 0.5153, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.3534577387486279, |
| "grad_norm": 9.258099865422958, |
| "learning_rate": 9.991388979491646e-07, |
| "loss": 0.4248, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.3556531284302964, |
| "grad_norm": 9.52300325877657, |
| "learning_rate": 9.990624156367596e-07, |
| "loss": 0.4744, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.3578485181119649, |
| "grad_norm": 8.87937529404168, |
| "learning_rate": 9.989826829134356e-07, |
| "loss": 0.4804, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.3600439077936334, |
| "grad_norm": 8.218898687145828, |
| "learning_rate": 9.988997002984949e-07, |
| "loss": 0.5154, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.36223929747530187, |
| "grad_norm": 9.404715926766505, |
| "learning_rate": 9.988134683324058e-07, |
| "loss": 0.4595, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.36443468715697036, |
| "grad_norm": 8.404212842327881, |
| "learning_rate": 9.987239875768006e-07, |
| "loss": 0.4122, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.36663007683863885, |
| "grad_norm": 12.927034279467245, |
| "learning_rate": 9.9863125861447e-07, |
| "loss": 0.458, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.36882546652030734, |
| "grad_norm": 9.395688075375638, |
| "learning_rate": 9.985352820493614e-07, |
| "loss": 0.4442, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.37102085620197583, |
| "grad_norm": 12.027605468945074, |
| "learning_rate": 9.984360585065733e-07, |
| "loss": 0.4342, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.3732162458836443, |
| "grad_norm": 11.101898867057535, |
| "learning_rate": 9.983335886323524e-07, |
| "loss": 0.4457, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.3754116355653128, |
| "grad_norm": 9.788751810099042, |
| "learning_rate": 9.98227873094088e-07, |
| "loss": 0.4773, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.37760702524698136, |
| "grad_norm": 8.772604827049461, |
| "learning_rate": 9.981189125803095e-07, |
| "loss": 0.4763, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.37980241492864986, |
| "grad_norm": 9.88650795305363, |
| "learning_rate": 9.980067078006804e-07, |
| "loss": 0.5062, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.38199780461031835, |
| "grad_norm": 8.73207633341288, |
| "learning_rate": 9.978912594859946e-07, |
| "loss": 0.5079, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.38419319429198684, |
| "grad_norm": 7.037929236057823, |
| "learning_rate": 9.977725683881707e-07, |
| "loss": 0.4444, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.38638858397365533, |
| "grad_norm": 14.22447093808163, |
| "learning_rate": 9.97650635280248e-07, |
| "loss": 0.4683, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.3885839736553238, |
| "grad_norm": 6.850018676344095, |
| "learning_rate": 9.97525460956381e-07, |
| "loss": 0.4544, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.3907793633369923, |
| "grad_norm": 9.295362171330144, |
| "learning_rate": 9.973970462318349e-07, |
| "loss": 0.4168, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.3929747530186608, |
| "grad_norm": 9.498462196078586, |
| "learning_rate": 9.972653919429788e-07, |
| "loss": 0.4762, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.3951701427003293, |
| "grad_norm": 14.51893007863248, |
| "learning_rate": 9.971304989472817e-07, |
| "loss": 0.429, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.3973655323819978, |
| "grad_norm": 10.861523669617597, |
| "learning_rate": 9.969923681233066e-07, |
| "loss": 0.5009, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.3995609220636663, |
| "grad_norm": 8.71310472180412, |
| "learning_rate": 9.968510003707042e-07, |
| "loss": 0.4562, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.40175631174533477, |
| "grad_norm": 7.504849700550564, |
| "learning_rate": 9.967063966102079e-07, |
| "loss": 0.4103, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.4039517014270033, |
| "grad_norm": 11.160809122945897, |
| "learning_rate": 9.965585577836264e-07, |
| "loss": 0.4868, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.4061470911086718, |
| "grad_norm": 9.173442546613508, |
| "learning_rate": 9.9640748485384e-07, |
| "loss": 0.4178, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.4083424807903403, |
| "grad_norm": 9.946380701067211, |
| "learning_rate": 9.962531788047913e-07, |
| "loss": 0.4075, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.4105378704720088, |
| "grad_norm": 6.7742539591322775, |
| "learning_rate": 9.960956406414813e-07, |
| "loss": 0.4228, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.4127332601536773, |
| "grad_norm": 6.667418828739606, |
| "learning_rate": 9.959348713899613e-07, |
| "loss": 0.4313, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.4149286498353458, |
| "grad_norm": 8.5318733439587, |
| "learning_rate": 9.957708720973273e-07, |
| "loss": 0.432, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.41712403951701427, |
| "grad_norm": 8.281669822513821, |
| "learning_rate": 9.956036438317123e-07, |
| "loss": 0.4454, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.41931942919868276, |
| "grad_norm": 11.094591963482126, |
| "learning_rate": 9.954331876822798e-07, |
| "loss": 0.4472, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.42151481888035125, |
| "grad_norm": 10.576004038338146, |
| "learning_rate": 9.952595047592167e-07, |
| "loss": 0.5284, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.42371020856201974, |
| "grad_norm": 7.579127144334551, |
| "learning_rate": 9.950825961937257e-07, |
| "loss": 0.4835, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.42590559824368823, |
| "grad_norm": 9.83385395859545, |
| "learning_rate": 9.949024631380189e-07, |
| "loss": 0.4513, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.4281009879253567, |
| "grad_norm": 9.260566516176205, |
| "learning_rate": 9.94719106765309e-07, |
| "loss": 0.4215, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.43029637760702527, |
| "grad_norm": 9.7492885786622, |
| "learning_rate": 9.945325282698022e-07, |
| "loss": 0.4338, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.43249176728869376, |
| "grad_norm": 8.278244359603962, |
| "learning_rate": 9.94342728866691e-07, |
| "loss": 0.4249, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.43468715697036225, |
| "grad_norm": 9.668030105849532, |
| "learning_rate": 9.941497097921456e-07, |
| "loss": 0.4328, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.43688254665203075, |
| "grad_norm": 9.271666205840873, |
| "learning_rate": 9.939534723033057e-07, |
| "loss": 0.4424, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.43907793633369924, |
| "grad_norm": 9.508865253140504, |
| "learning_rate": 9.937540176782731e-07, |
| "loss": 0.4076, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.43907793633369924, |
| "eval_accuracy": 0.778, |
| "eval_loss": 0.41148170828819275, |
| "eval_runtime": 52.0039, |
| "eval_samples_per_second": 9.615, |
| "eval_steps_per_second": 1.211, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.44127332601536773, |
| "grad_norm": 15.641606615350042, |
| "learning_rate": 9.935513472161026e-07, |
| "loss": 0.5175, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.4434687156970362, |
| "grad_norm": 6.583946691921041, |
| "learning_rate": 9.93345462236794e-07, |
| "loss": 0.4612, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.4456641053787047, |
| "grad_norm": 7.317576490413668, |
| "learning_rate": 9.931363640812837e-07, |
| "loss": 0.4319, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.4478594950603732, |
| "grad_norm": 8.58238734593965, |
| "learning_rate": 9.929240541114347e-07, |
| "loss": 0.4362, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.4500548847420417, |
| "grad_norm": 7.645878462772526, |
| "learning_rate": 9.927085337100298e-07, |
| "loss": 0.4449, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.4522502744237102, |
| "grad_norm": 8.128944249572761, |
| "learning_rate": 9.924898042807604e-07, |
| "loss": 0.4359, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.4544456641053787, |
| "grad_norm": 6.323254265724658, |
| "learning_rate": 9.922678672482192e-07, |
| "loss": 0.4349, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.4566410537870472, |
| "grad_norm": 8.343466628765478, |
| "learning_rate": 9.920427240578898e-07, |
| "loss": 0.4601, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.4588364434687157, |
| "grad_norm": 9.94252397747147, |
| "learning_rate": 9.918143761761376e-07, |
| "loss": 0.4275, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.4610318331503842, |
| "grad_norm": 6.996773271668393, |
| "learning_rate": 9.915828250902003e-07, |
| "loss": 0.4092, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.4632272228320527, |
| "grad_norm": 7.702375964154107, |
| "learning_rate": 9.913480723081782e-07, |
| "loss": 0.4039, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.4654226125137212, |
| "grad_norm": 12.275736080317706, |
| "learning_rate": 9.911101193590243e-07, |
| "loss": 0.483, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.4676180021953897, |
| "grad_norm": 13.554222149681413, |
| "learning_rate": 9.908689677925347e-07, |
| "loss": 0.408, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.4698133918770582, |
| "grad_norm": 10.859860574414233, |
| "learning_rate": 9.906246191793378e-07, |
| "loss": 0.4032, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.47200878155872666, |
| "grad_norm": 13.78109413813406, |
| "learning_rate": 9.903770751108845e-07, |
| "loss": 0.414, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.47420417124039516, |
| "grad_norm": 7.982093750602254, |
| "learning_rate": 9.901263371994381e-07, |
| "loss": 0.4097, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.47639956092206365, |
| "grad_norm": 7.226081741797403, |
| "learning_rate": 9.898724070780636e-07, |
| "loss": 0.424, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.47859495060373214, |
| "grad_norm": 6.5505528789866005, |
| "learning_rate": 9.896152864006163e-07, |
| "loss": 0.4117, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.4807903402854007, |
| "grad_norm": 7.238329656366918, |
| "learning_rate": 9.893549768417324e-07, |
| "loss": 0.426, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.4829857299670692, |
| "grad_norm": 9.018371675867607, |
| "learning_rate": 9.89091480096817e-07, |
| "loss": 0.4301, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.48518111964873767, |
| "grad_norm": 6.873575031102029, |
| "learning_rate": 9.888247978820336e-07, |
| "loss": 0.4412, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.48737650933040616, |
| "grad_norm": 7.71741695116787, |
| "learning_rate": 9.88554931934293e-07, |
| "loss": 0.401, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.48957189901207465, |
| "grad_norm": 6.450074931709192, |
| "learning_rate": 9.882818840112412e-07, |
| "loss": 0.4453, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.49176728869374314, |
| "grad_norm": 10.18694339380834, |
| "learning_rate": 9.88005655891249e-07, |
| "loss": 0.4499, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.49396267837541163, |
| "grad_norm": 8.526454527609461, |
| "learning_rate": 9.877262493734e-07, |
| "loss": 0.3529, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.4961580680570801, |
| "grad_norm": 7.8436191187666795, |
| "learning_rate": 9.874436662774781e-07, |
| "loss": 0.3862, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.4983534577387486, |
| "grad_norm": 12.16130920239228, |
| "learning_rate": 9.871579084439573e-07, |
| "loss": 0.4494, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.5005488474204172, |
| "grad_norm": 13.22347671983063, |
| "learning_rate": 9.868689777339882e-07, |
| "loss": 0.3777, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.5027442371020856, |
| "grad_norm": 10.714878837498341, |
| "learning_rate": 9.865768760293865e-07, |
| "loss": 0.4392, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.5049396267837541, |
| "grad_norm": 10.989931098186185, |
| "learning_rate": 9.862816052326207e-07, |
| "loss": 0.4342, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.5071350164654226, |
| "grad_norm": 9.420235479586577, |
| "learning_rate": 9.859831672668001e-07, |
| "loss": 0.449, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.5093304061470911, |
| "grad_norm": 7.8352789241534895, |
| "learning_rate": 9.856815640756614e-07, |
| "loss": 0.4066, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.5115257958287596, |
| "grad_norm": 6.528093022715006, |
| "learning_rate": 9.85376797623557e-07, |
| "loss": 0.4076, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.5137211855104281, |
| "grad_norm": 9.599510824703753, |
| "learning_rate": 9.850688698954408e-07, |
| "loss": 0.4656, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.5159165751920965, |
| "grad_norm": 7.327102993121418, |
| "learning_rate": 9.847577828968574e-07, |
| "loss": 0.3928, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.5181119648737651, |
| "grad_norm": 8.47025842420312, |
| "learning_rate": 9.84443538653927e-07, |
| "loss": 0.4459, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.5203073545554336, |
| "grad_norm": 8.808300848288402, |
| "learning_rate": 9.841261392133334e-07, |
| "loss": 0.4699, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.5225027442371021, |
| "grad_norm": 6.4472715749678535, |
| "learning_rate": 9.838055866423101e-07, |
| "loss": 0.3927, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.5246981339187706, |
| "grad_norm": 8.71789150223208, |
| "learning_rate": 9.834818830286274e-07, |
| "loss": 0.4112, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.5268935236004391, |
| "grad_norm": 6.520504611614625, |
| "learning_rate": 9.83155030480578e-07, |
| "loss": 0.4221, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.5290889132821076, |
| "grad_norm": 7.822762966760966, |
| "learning_rate": 9.82825031126964e-07, |
| "loss": 0.4085, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.531284302963776, |
| "grad_norm": 8.42042192814293, |
| "learning_rate": 9.82491887117083e-07, |
| "loss": 0.3344, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.5334796926454446, |
| "grad_norm": 8.795849170631488, |
| "learning_rate": 9.821556006207131e-07, |
| "loss": 0.4706, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.535675082327113, |
| "grad_norm": 11.237650950766433, |
| "learning_rate": 9.818161738281003e-07, |
| "loss": 0.3922, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.5378704720087816, |
| "grad_norm": 8.79761175477303, |
| "learning_rate": 9.81473608949943e-07, |
| "loss": 0.3798, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.54006586169045, |
| "grad_norm": 9.224818726295643, |
| "learning_rate": 9.811279082173783e-07, |
| "loss": 0.4045, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.5422612513721186, |
| "grad_norm": 11.451420498389481, |
| "learning_rate": 9.80779073881967e-07, |
| "loss": 0.4767, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.544456641053787, |
| "grad_norm": 14.204545837436886, |
| "learning_rate": 9.804271082156792e-07, |
| "loss": 0.5448, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.5466520307354555, |
| "grad_norm": 8.162056177470248, |
| "learning_rate": 9.800720135108798e-07, |
| "loss": 0.4662, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.5488474204171241, |
| "grad_norm": 5.928770792444553, |
| "learning_rate": 9.79713792080313e-07, |
| "loss": 0.3764, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.5488474204171241, |
| "eval_accuracy": 0.764, |
| "eval_loss": 0.41615644097328186, |
| "eval_runtime": 51.876, |
| "eval_samples_per_second": 9.638, |
| "eval_steps_per_second": 1.214, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.5510428100987925, |
| "grad_norm": 6.184202969509846, |
| "learning_rate": 9.793524462570874e-07, |
| "loss": 0.3895, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.5532381997804611, |
| "grad_norm": 7.749574593570711, |
| "learning_rate": 9.78987978394661e-07, |
| "loss": 0.4515, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.5554335894621295, |
| "grad_norm": 8.854833970278174, |
| "learning_rate": 9.786203908668255e-07, |
| "loss": 0.4269, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.557628979143798, |
| "grad_norm": 6.18988995877324, |
| "learning_rate": 9.78249686067691e-07, |
| "loss": 0.4204, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.5598243688254665, |
| "grad_norm": 6.0975770712127, |
| "learning_rate": 9.778758664116717e-07, |
| "loss": 0.4312, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.562019758507135, |
| "grad_norm": 7.860127371723332, |
| "learning_rate": 9.774989343334675e-07, |
| "loss": 0.4056, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.5642151481888035, |
| "grad_norm": 6.287357081202248, |
| "learning_rate": 9.771188922880501e-07, |
| "loss": 0.431, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.566410537870472, |
| "grad_norm": 7.482025843040615, |
| "learning_rate": 9.76735742750647e-07, |
| "loss": 0.4237, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.5686059275521405, |
| "grad_norm": 6.626343732308425, |
| "learning_rate": 9.763494882167238e-07, |
| "loss": 0.4258, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.570801317233809, |
| "grad_norm": 6.379142997039637, |
| "learning_rate": 9.759601312019705e-07, |
| "loss": 0.4496, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.5729967069154775, |
| "grad_norm": 7.786859033495975, |
| "learning_rate": 9.755676742422824e-07, |
| "loss": 0.3697, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.575192096597146, |
| "grad_norm": 7.474508144431387, |
| "learning_rate": 9.751721198937457e-07, |
| "loss": 0.3805, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.5773874862788145, |
| "grad_norm": 5.797750979716916, |
| "learning_rate": 9.747734707326194e-07, |
| "loss": 0.3832, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.579582875960483, |
| "grad_norm": 10.801633788158538, |
| "learning_rate": 9.743717293553197e-07, |
| "loss": 0.3489, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.5817782656421515, |
| "grad_norm": 8.81184130305926, |
| "learning_rate": 9.73966898378402e-07, |
| "loss": 0.3933, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.58397365532382, |
| "grad_norm": 10.99886571046786, |
| "learning_rate": 9.735589804385445e-07, |
| "loss": 0.4582, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.5861690450054885, |
| "grad_norm": 8.989645189656823, |
| "learning_rate": 9.731479781925308e-07, |
| "loss": 0.39, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.5883644346871569, |
| "grad_norm": 9.04325471249987, |
| "learning_rate": 9.727338943172335e-07, |
| "loss": 0.402, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.5905598243688255, |
| "grad_norm": 8.32075801154753, |
| "learning_rate": 9.723167315095947e-07, |
| "loss": 0.3896, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.5927552140504939, |
| "grad_norm": 11.735779106191535, |
| "learning_rate": 9.718964924866108e-07, |
| "loss": 0.4977, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.5949506037321625, |
| "grad_norm": 6.9875015531767515, |
| "learning_rate": 9.71473179985313e-07, |
| "loss": 0.4142, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.5971459934138309, |
| "grad_norm": 7.256558656488309, |
| "learning_rate": 9.710467967627502e-07, |
| "loss": 0.4729, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.5993413830954994, |
| "grad_norm": 6.63170478635617, |
| "learning_rate": 9.706173455959713e-07, |
| "loss": 0.3926, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.601536772777168, |
| "grad_norm": 7.736941306994274, |
| "learning_rate": 9.701848292820069e-07, |
| "loss": 0.4269, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.6037321624588364, |
| "grad_norm": 7.141227825990968, |
| "learning_rate": 9.697492506378507e-07, |
| "loss": 0.4746, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.605927552140505, |
| "grad_norm": 6.175085064537569, |
| "learning_rate": 9.693106125004416e-07, |
| "loss": 0.3916, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.6081229418221734, |
| "grad_norm": 8.496732166784003, |
| "learning_rate": 9.688689177266452e-07, |
| "loss": 0.4195, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.610318331503842, |
| "grad_norm": 5.940495748803936, |
| "learning_rate": 9.684241691932347e-07, |
| "loss": 0.4218, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.6125137211855104, |
| "grad_norm": 6.794809207027829, |
| "learning_rate": 9.679763697968732e-07, |
| "loss": 0.4064, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.6147091108671789, |
| "grad_norm": 7.232988324397075, |
| "learning_rate": 9.675255224540934e-07, |
| "loss": 0.3699, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.6169045005488474, |
| "grad_norm": 7.07939070544042, |
| "learning_rate": 9.6707163010128e-07, |
| "loss": 0.4168, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.6190998902305159, |
| "grad_norm": 8.718610382883304, |
| "learning_rate": 9.666146956946496e-07, |
| "loss": 0.4097, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.6212952799121844, |
| "grad_norm": 9.093471494946654, |
| "learning_rate": 9.661547222102321e-07, |
| "loss": 0.4252, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.6234906695938529, |
| "grad_norm": 13.940230416519052, |
| "learning_rate": 9.656917126438508e-07, |
| "loss": 0.4567, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.6256860592755215, |
| "grad_norm": 8.779803873571954, |
| "learning_rate": 9.65225670011103e-07, |
| "loss": 0.383, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.6278814489571899, |
| "grad_norm": 7.5708107372450035, |
| "learning_rate": 9.647565973473407e-07, |
| "loss": 0.4124, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.6300768386388584, |
| "grad_norm": 13.174216612702878, |
| "learning_rate": 9.642844977076507e-07, |
| "loss": 0.4502, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.6322722283205269, |
| "grad_norm": 8.49295896068912, |
| "learning_rate": 9.63809374166834e-07, |
| "loss": 0.4369, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.6344676180021954, |
| "grad_norm": 9.345508533381187, |
| "learning_rate": 9.633312298193871e-07, |
| "loss": 0.4625, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.6366630076838639, |
| "grad_norm": 9.696466925170553, |
| "learning_rate": 9.62850067779481e-07, |
| "loss": 0.4442, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.6388583973655324, |
| "grad_norm": 8.896625991218926, |
| "learning_rate": 9.623658911809404e-07, |
| "loss": 0.3991, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.6410537870472008, |
| "grad_norm": 9.525265923938584, |
| "learning_rate": 9.618787031772245e-07, |
| "loss": 0.3849, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.6432491767288694, |
| "grad_norm": 5.576065012637179, |
| "learning_rate": 9.61388506941406e-07, |
| "loss": 0.4233, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.6454445664105378, |
| "grad_norm": 7.207790760174152, |
| "learning_rate": 9.6089530566615e-07, |
| "loss": 0.3935, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.6476399560922064, |
| "grad_norm": 6.129075752774988, |
| "learning_rate": 9.603991025636933e-07, |
| "loss": 0.3904, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.6498353457738749, |
| "grad_norm": 7.223439076532464, |
| "learning_rate": 9.598999008658241e-07, |
| "loss": 0.3842, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.6520307354555434, |
| "grad_norm": 8.143383508497742, |
| "learning_rate": 9.59397703823861e-07, |
| "loss": 0.4269, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.6542261251372119, |
| "grad_norm": 9.02225167943478, |
| "learning_rate": 9.588925147086303e-07, |
| "loss": 0.3878, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.6564215148188803, |
| "grad_norm": 7.737039238378963, |
| "learning_rate": 9.583843368104464e-07, |
| "loss": 0.3508, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.6586169045005489, |
| "grad_norm": 6.743123024533224, |
| "learning_rate": 9.578731734390898e-07, |
| "loss": 0.3808, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.6586169045005489, |
| "eval_accuracy": 0.77, |
| "eval_loss": 0.41097283363342285, |
| "eval_runtime": 52.0828, |
| "eval_samples_per_second": 9.6, |
| "eval_steps_per_second": 1.21, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.6608122941822173, |
| "grad_norm": 7.137538658423445, |
| "learning_rate": 9.573590279237854e-07, |
| "loss": 0.3938, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.6630076838638859, |
| "grad_norm": 6.92894058529831, |
| "learning_rate": 9.568419036131807e-07, |
| "loss": 0.3729, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.6652030735455543, |
| "grad_norm": 8.370178365332173, |
| "learning_rate": 9.563218038753245e-07, |
| "loss": 0.4004, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.6673984632272228, |
| "grad_norm": 14.386262804761378, |
| "learning_rate": 9.557987320976446e-07, |
| "loss": 0.4211, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.6695938529088913, |
| "grad_norm": 7.7875110721670975, |
| "learning_rate": 9.552726916869254e-07, |
| "loss": 0.388, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.6717892425905598, |
| "grad_norm": 8.49964853234254, |
| "learning_rate": 9.547436860692869e-07, |
| "loss": 0.4369, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.6739846322722283, |
| "grad_norm": 6.5964314472495404, |
| "learning_rate": 9.542117186901608e-07, |
| "loss": 0.3423, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.6761800219538968, |
| "grad_norm": 7.6505759115431085, |
| "learning_rate": 9.536767930142692e-07, |
| "loss": 0.3602, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.6783754116355654, |
| "grad_norm": 7.556175808157708, |
| "learning_rate": 9.53138912525602e-07, |
| "loss": 0.4385, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.6805708013172338, |
| "grad_norm": 9.165395898058962, |
| "learning_rate": 9.525980807273933e-07, |
| "loss": 0.4516, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.6827661909989023, |
| "grad_norm": 7.637449593962663, |
| "learning_rate": 9.520543011420994e-07, |
| "loss": 0.4289, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.6849615806805708, |
| "grad_norm": 9.18857916606884, |
| "learning_rate": 9.515075773113758e-07, |
| "loss": 0.3785, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.6871569703622393, |
| "grad_norm": 7.821219676613859, |
| "learning_rate": 9.509579127960541e-07, |
| "loss": 0.4158, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.6893523600439078, |
| "grad_norm": 6.571343391685928, |
| "learning_rate": 9.504053111761183e-07, |
| "loss": 0.3609, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.6915477497255763, |
| "grad_norm": 6.225595749149676, |
| "learning_rate": 9.498497760506819e-07, |
| "loss": 0.4359, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.6937431394072447, |
| "grad_norm": 6.950774628316192, |
| "learning_rate": 9.492913110379647e-07, |
| "loss": 0.4417, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.6959385290889133, |
| "grad_norm": 5.612528566630172, |
| "learning_rate": 9.487299197752687e-07, |
| "loss": 0.4052, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.6981339187705817, |
| "grad_norm": 8.404168783130382, |
| "learning_rate": 9.481656059189549e-07, |
| "loss": 0.4162, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.7003293084522503, |
| "grad_norm": 8.031865429035927, |
| "learning_rate": 9.475983731444191e-07, |
| "loss": 0.4125, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.7025246981339188, |
| "grad_norm": 10.321047597755841, |
| "learning_rate": 9.47028225146068e-07, |
| "loss": 0.4505, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.7047200878155873, |
| "grad_norm": 7.740999393168428, |
| "learning_rate": 9.464551656372955e-07, |
| "loss": 0.3765, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.7069154774972558, |
| "grad_norm": 8.215346652475102, |
| "learning_rate": 9.458791983504581e-07, |
| "loss": 0.401, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.7091108671789242, |
| "grad_norm": 7.4542945597526336, |
| "learning_rate": 9.453003270368509e-07, |
| "loss": 0.4796, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.7113062568605928, |
| "grad_norm": 7.054708809148231, |
| "learning_rate": 9.44718555466683e-07, |
| "loss": 0.4497, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.7135016465422612, |
| "grad_norm": 7.460280928904919, |
| "learning_rate": 9.44133887429053e-07, |
| "loss": 0.3926, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.7156970362239298, |
| "grad_norm": 8.265245053890029, |
| "learning_rate": 9.435463267319239e-07, |
| "loss": 0.3805, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.7178924259055982, |
| "grad_norm": 7.800641885443644, |
| "learning_rate": 9.429558772020992e-07, |
| "loss": 0.4424, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.7200878155872668, |
| "grad_norm": 8.978144769546304, |
| "learning_rate": 9.423625426851973e-07, |
| "loss": 0.4159, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.7222832052689352, |
| "grad_norm": 6.867455347578303, |
| "learning_rate": 9.417663270456267e-07, |
| "loss": 0.402, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.7244785949506037, |
| "grad_norm": 7.15816324078702, |
| "learning_rate": 9.411672341665604e-07, |
| "loss": 0.3814, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.7266739846322722, |
| "grad_norm": 7.444204700173956, |
| "learning_rate": 9.405652679499115e-07, |
| "loss": 0.3782, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.7288693743139407, |
| "grad_norm": 6.789459839990034, |
| "learning_rate": 9.399604323163068e-07, |
| "loss": 0.3484, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.7310647639956093, |
| "grad_norm": 14.478524238054472, |
| "learning_rate": 9.393527312050617e-07, |
| "loss": 0.464, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.7332601536772777, |
| "grad_norm": 20.668512412883626, |
| "learning_rate": 9.387421685741552e-07, |
| "loss": 0.4503, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.7354555433589463, |
| "grad_norm": 12.197800422910193, |
| "learning_rate": 9.381287484002027e-07, |
| "loss": 0.4281, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.7376509330406147, |
| "grad_norm": 6.445277090460248, |
| "learning_rate": 9.375124746784311e-07, |
| "loss": 0.3648, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.7398463227222832, |
| "grad_norm": 11.96859661106773, |
| "learning_rate": 9.368933514226529e-07, |
| "loss": 0.4608, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.7420417124039517, |
| "grad_norm": 5.939377322140176, |
| "learning_rate": 9.362713826652392e-07, |
| "loss": 0.4006, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.7442371020856202, |
| "grad_norm": 8.561222995237147, |
| "learning_rate": 9.356465724570943e-07, |
| "loss": 0.4163, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.7464324917672887, |
| "grad_norm": 6.033330356983505, |
| "learning_rate": 9.350189248676292e-07, |
| "loss": 0.4144, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.7486278814489572, |
| "grad_norm": 6.205787168601971, |
| "learning_rate": 9.34388443984734e-07, |
| "loss": 0.4258, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.7508232711306256, |
| "grad_norm": 7.631254905556204, |
| "learning_rate": 9.33755133914753e-07, |
| "loss": 0.4065, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.7530186608122942, |
| "grad_norm": 6.32445637747235, |
| "learning_rate": 9.331189987824568e-07, |
| "loss": 0.4389, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.7552140504939627, |
| "grad_norm": 5.65488048442344, |
| "learning_rate": 9.324800427310155e-07, |
| "loss": 0.4363, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.7574094401756312, |
| "grad_norm": 6.166240836463136, |
| "learning_rate": 9.318382699219722e-07, |
| "loss": 0.4091, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.7596048298572997, |
| "grad_norm": 7.489797283268898, |
| "learning_rate": 9.311936845352157e-07, |
| "loss": 0.3923, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.7618002195389681, |
| "grad_norm": 7.098266537437995, |
| "learning_rate": 9.305462907689532e-07, |
| "loss": 0.4363, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.7639956092206367, |
| "grad_norm": 7.518882790839629, |
| "learning_rate": 9.298960928396826e-07, |
| "loss": 0.3989, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.7661909989023051, |
| "grad_norm": 7.535014992332507, |
| "learning_rate": 9.292430949821659e-07, |
| "loss": 0.3939, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.7683863885839737, |
| "grad_norm": 7.793847428801205, |
| "learning_rate": 9.285873014494008e-07, |
| "loss": 0.4238, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.7683863885839737, |
| "eval_accuracy": 0.78, |
| "eval_loss": 0.3999713063240051, |
| "eval_runtime": 52.2328, |
| "eval_samples_per_second": 9.573, |
| "eval_steps_per_second": 1.206, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.7705817782656421, |
| "grad_norm": 6.9078346633967405, |
| "learning_rate": 9.279287165125936e-07, |
| "loss": 0.3417, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.7727771679473107, |
| "grad_norm": 8.452074250696336, |
| "learning_rate": 9.272673444611308e-07, |
| "loss": 0.426, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.7749725576289791, |
| "grad_norm": 6.7224178173856215, |
| "learning_rate": 9.266031896025516e-07, |
| "loss": 0.421, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.7771679473106476, |
| "grad_norm": 8.735363673640215, |
| "learning_rate": 9.259362562625199e-07, |
| "loss": 0.3596, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.7793633369923162, |
| "grad_norm": 7.949892246537311, |
| "learning_rate": 9.252665487847957e-07, |
| "loss": 0.3922, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.7815587266739846, |
| "grad_norm": 12.194367861210461, |
| "learning_rate": 9.245940715312074e-07, |
| "loss": 0.4538, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.7837541163556532, |
| "grad_norm": 10.45987571002838, |
| "learning_rate": 9.239188288816226e-07, |
| "loss": 0.4928, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.7859495060373216, |
| "grad_norm": 6.764374108633695, |
| "learning_rate": 9.232408252339201e-07, |
| "loss": 0.4226, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.7881448957189902, |
| "grad_norm": 6.394843611556007, |
| "learning_rate": 9.225600650039615e-07, |
| "loss": 0.4096, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.7903402854006586, |
| "grad_norm": 6.971937381598142, |
| "learning_rate": 9.218765526255619e-07, |
| "loss": 0.4162, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.7925356750823271, |
| "grad_norm": 5.9443940673499185, |
| "learning_rate": 9.211902925504613e-07, |
| "loss": 0.4161, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.7947310647639956, |
| "grad_norm": 5.483471181116708, |
| "learning_rate": 9.205012892482952e-07, |
| "loss": 0.3627, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.7969264544456641, |
| "grad_norm": 6.649198037257637, |
| "learning_rate": 9.198095472065667e-07, |
| "loss": 0.445, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.7991218441273326, |
| "grad_norm": 7.566125395640005, |
| "learning_rate": 9.191150709306155e-07, |
| "loss": 0.4352, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.8013172338090011, |
| "grad_norm": 5.877051153754724, |
| "learning_rate": 9.184178649435896e-07, |
| "loss": 0.3855, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.8035126234906695, |
| "grad_norm": 6.82408167677956, |
| "learning_rate": 9.177179337864163e-07, |
| "loss": 0.44, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.8057080131723381, |
| "grad_norm": 5.76805550171372, |
| "learning_rate": 9.170152820177714e-07, |
| "loss": 0.3722, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.8079034028540066, |
| "grad_norm": 5.37322049457271, |
| "learning_rate": 9.163099142140505e-07, |
| "loss": 0.39, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.8100987925356751, |
| "grad_norm": 6.652734325876614, |
| "learning_rate": 9.156018349693386e-07, |
| "loss": 0.33, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.8122941822173436, |
| "grad_norm": 8.078887258485842, |
| "learning_rate": 9.148910488953807e-07, |
| "loss": 0.4495, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.814489571899012, |
| "grad_norm": 6.057501262886995, |
| "learning_rate": 9.141775606215512e-07, |
| "loss": 0.3793, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.8166849615806806, |
| "grad_norm": 8.349159344050063, |
| "learning_rate": 9.134613747948238e-07, |
| "loss": 0.4165, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.818880351262349, |
| "grad_norm": 6.093265741529057, |
| "learning_rate": 9.127424960797423e-07, |
| "loss": 0.3683, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.8210757409440176, |
| "grad_norm": 7.626403918604381, |
| "learning_rate": 9.120209291583885e-07, |
| "loss": 0.4478, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.823271130625686, |
| "grad_norm": 6.286457043534019, |
| "learning_rate": 9.11296678730353e-07, |
| "loss": 0.4432, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.8254665203073546, |
| "grad_norm": 6.1102728691763195, |
| "learning_rate": 9.10569749512704e-07, |
| "loss": 0.4408, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.827661909989023, |
| "grad_norm": 5.8435296710540765, |
| "learning_rate": 9.098401462399572e-07, |
| "loss": 0.3889, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.8298572996706916, |
| "grad_norm": 6.8221898822791776, |
| "learning_rate": 9.091078736640438e-07, |
| "loss": 0.3924, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.8320526893523601, |
| "grad_norm": 5.449250972840727, |
| "learning_rate": 9.083729365542807e-07, |
| "loss": 0.4544, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.8342480790340285, |
| "grad_norm": 6.531068732007626, |
| "learning_rate": 9.076353396973391e-07, |
| "loss": 0.3824, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.8364434687156971, |
| "grad_norm": 6.253901506150054, |
| "learning_rate": 9.068950878972128e-07, |
| "loss": 0.393, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.8386388583973655, |
| "grad_norm": 6.104772949983095, |
| "learning_rate": 9.06152185975188e-07, |
| "loss": 0.3857, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.8408342480790341, |
| "grad_norm": 5.459509353786465, |
| "learning_rate": 9.054066387698103e-07, |
| "loss": 0.3748, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.8430296377607025, |
| "grad_norm": 6.221629121824156, |
| "learning_rate": 9.04658451136855e-07, |
| "loss": 0.3869, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.845225027442371, |
| "grad_norm": 11.117787770421263, |
| "learning_rate": 9.039076279492938e-07, |
| "loss": 0.428, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.8474204171240395, |
| "grad_norm": 7.6402881827025695, |
| "learning_rate": 9.03154174097265e-07, |
| "loss": 0.3685, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.849615806805708, |
| "grad_norm": 8.329862332560323, |
| "learning_rate": 9.023980944880395e-07, |
| "loss": 0.3888, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.8518111964873765, |
| "grad_norm": 9.379807197876737, |
| "learning_rate": 9.016393940459901e-07, |
| "loss": 0.3228, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.854006586169045, |
| "grad_norm": 9.497050528776587, |
| "learning_rate": 9.008780777125592e-07, |
| "loss": 0.3643, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.8562019758507134, |
| "grad_norm": 11.317540635350868, |
| "learning_rate": 9.001141504462267e-07, |
| "loss": 0.3849, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.858397365532382, |
| "grad_norm": 9.649076333363197, |
| "learning_rate": 8.993476172224776e-07, |
| "loss": 0.4216, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.8605927552140505, |
| "grad_norm": 5.870968468017432, |
| "learning_rate": 8.985784830337694e-07, |
| "loss": 0.3512, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.862788144895719, |
| "grad_norm": 6.63746206403619, |
| "learning_rate": 8.978067528895001e-07, |
| "loss": 0.4034, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.8649835345773875, |
| "grad_norm": 7.9614797293095805, |
| "learning_rate": 8.970324318159747e-07, |
| "loss": 0.4374, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.867178924259056, |
| "grad_norm": 6.6259673033312305, |
| "learning_rate": 8.962555248563737e-07, |
| "loss": 0.4034, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.8693743139407245, |
| "grad_norm": 5.874087728977652, |
| "learning_rate": 8.95476037070719e-07, |
| "loss": 0.4381, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.8715697036223929, |
| "grad_norm": 5.374289997565401, |
| "learning_rate": 8.94693973535842e-07, |
| "loss": 0.4015, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.8737650933040615, |
| "grad_norm": 5.135571917512859, |
| "learning_rate": 8.939093393453494e-07, |
| "loss": 0.3764, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.8759604829857299, |
| "grad_norm": 6.320197335683808, |
| "learning_rate": 8.931221396095914e-07, |
| "loss": 0.3573, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.8781558726673985, |
| "grad_norm": 5.934124946312254, |
| "learning_rate": 8.92332379455627e-07, |
| "loss": 0.3985, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.8781558726673985, |
| "eval_accuracy": 0.78, |
| "eval_loss": 0.3908381462097168, |
| "eval_runtime": 52.2922, |
| "eval_samples_per_second": 9.562, |
| "eval_steps_per_second": 1.205, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.8803512623490669, |
| "grad_norm": 5.227662807462205, |
| "learning_rate": 8.91540064027192e-07, |
| "loss": 0.3802, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.8825466520307355, |
| "grad_norm": 5.4312238244768105, |
| "learning_rate": 8.907451984846642e-07, |
| "loss": 0.357, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.884742041712404, |
| "grad_norm": 6.5227729333700255, |
| "learning_rate": 8.899477880050305e-07, |
| "loss": 0.4599, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.8869374313940724, |
| "grad_norm": 6.320159868176597, |
| "learning_rate": 8.891478377818533e-07, |
| "loss": 0.3456, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.889132821075741, |
| "grad_norm": 6.595230869297455, |
| "learning_rate": 8.883453530252363e-07, |
| "loss": 0.3385, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.8913282107574094, |
| "grad_norm": 8.372111940632312, |
| "learning_rate": 8.875403389617909e-07, |
| "loss": 0.4228, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.893523600439078, |
| "grad_norm": 7.273966750571457, |
| "learning_rate": 8.867328008346012e-07, |
| "loss": 0.378, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.8957189901207464, |
| "grad_norm": 9.685872652254469, |
| "learning_rate": 8.859227439031917e-07, |
| "loss": 0.4191, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.897914379802415, |
| "grad_norm": 9.466266479949322, |
| "learning_rate": 8.851101734434916e-07, |
| "loss": 0.3949, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.9001097694840834, |
| "grad_norm": 8.975473353994817, |
| "learning_rate": 8.842950947478001e-07, |
| "loss": 0.4137, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.9023051591657519, |
| "grad_norm": 10.051513089446058, |
| "learning_rate": 8.834775131247534e-07, |
| "loss": 0.4055, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.9045005488474204, |
| "grad_norm": 7.417826600936023, |
| "learning_rate": 8.826574338992893e-07, |
| "loss": 0.4014, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.9066959385290889, |
| "grad_norm": 7.7383905505396475, |
| "learning_rate": 8.818348624126122e-07, |
| "loss": 0.3484, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.9088913282107574, |
| "grad_norm": 7.082975830005012, |
| "learning_rate": 8.810098040221588e-07, |
| "loss": 0.3709, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.9110867178924259, |
| "grad_norm": 6.727376213025829, |
| "learning_rate": 8.801822641015635e-07, |
| "loss": 0.3843, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.9132821075740944, |
| "grad_norm": 7.275555271868482, |
| "learning_rate": 8.793522480406223e-07, |
| "loss": 0.356, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.9154774972557629, |
| "grad_norm": 5.944537715719952, |
| "learning_rate": 8.785197612452591e-07, |
| "loss": 0.4293, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.9176728869374314, |
| "grad_norm": 6.472465668641822, |
| "learning_rate": 8.776848091374892e-07, |
| "loss": 0.353, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.9198682766190999, |
| "grad_norm": 7.887473362612184, |
| "learning_rate": 8.768473971553847e-07, |
| "loss": 0.417, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.9220636663007684, |
| "grad_norm": 5.72620013047636, |
| "learning_rate": 8.760075307530392e-07, |
| "loss": 0.3725, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.9242590559824369, |
| "grad_norm": 6.518816397700794, |
| "learning_rate": 8.75165215400532e-07, |
| "loss": 0.3819, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.9264544456641054, |
| "grad_norm": 7.2515922225795695, |
| "learning_rate": 8.743204565838922e-07, |
| "loss": 0.4082, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.9286498353457738, |
| "grad_norm": 6.156643476702157, |
| "learning_rate": 8.734732598050636e-07, |
| "loss": 0.4111, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.9308452250274424, |
| "grad_norm": 6.1131283071130635, |
| "learning_rate": 8.726236305818681e-07, |
| "loss": 0.3849, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.9330406147091108, |
| "grad_norm": 8.044334742394684, |
| "learning_rate": 8.717715744479706e-07, |
| "loss": 0.4006, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.9352360043907794, |
| "grad_norm": 6.327277854929185, |
| "learning_rate": 8.709170969528425e-07, |
| "loss": 0.4525, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.9374313940724479, |
| "grad_norm": 6.919003187239627, |
| "learning_rate": 8.700602036617253e-07, |
| "loss": 0.3976, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.9396267837541163, |
| "grad_norm": 6.750362117722639, |
| "learning_rate": 8.692009001555951e-07, |
| "loss": 0.4214, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.9418221734357849, |
| "grad_norm": 5.309527010131869, |
| "learning_rate": 8.683391920311256e-07, |
| "loss": 0.3837, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.9440175631174533, |
| "grad_norm": 5.756043416124202, |
| "learning_rate": 8.674750849006518e-07, |
| "loss": 0.3748, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.9462129527991219, |
| "grad_norm": 5.947118773957151, |
| "learning_rate": 8.666085843921337e-07, |
| "loss": 0.3656, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.9484083424807903, |
| "grad_norm": 6.803175272777888, |
| "learning_rate": 8.65739696149119e-07, |
| "loss": 0.3871, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.9506037321624589, |
| "grad_norm": 11.872613980622894, |
| "learning_rate": 8.648684258307075e-07, |
| "loss": 0.4402, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.9527991218441273, |
| "grad_norm": 11.845458798682339, |
| "learning_rate": 8.639947791115131e-07, |
| "loss": 0.398, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.9549945115257958, |
| "grad_norm": 7.885958732125603, |
| "learning_rate": 8.631187616816271e-07, |
| "loss": 0.3649, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.9571899012074643, |
| "grad_norm": 7.087463703257775, |
| "learning_rate": 8.622403792465819e-07, |
| "loss": 0.3938, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.9593852908891328, |
| "grad_norm": 6.677319825795207, |
| "learning_rate": 8.613596375273127e-07, |
| "loss": 0.379, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.9615806805708014, |
| "grad_norm": 5.657521052821894, |
| "learning_rate": 8.604765422601213e-07, |
| "loss": 0.3482, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.9637760702524698, |
| "grad_norm": 5.630660484188195, |
| "learning_rate": 8.595910991966375e-07, |
| "loss": 0.4039, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.9659714599341384, |
| "grad_norm": 6.274352234189354, |
| "learning_rate": 8.587033141037833e-07, |
| "loss": 0.3926, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.9681668496158068, |
| "grad_norm": 6.263591446228187, |
| "learning_rate": 8.578131927637339e-07, |
| "loss": 0.3528, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.9703622392974753, |
| "grad_norm": 7.115149701513277, |
| "learning_rate": 8.569207409738804e-07, |
| "loss": 0.3812, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.9725576289791438, |
| "grad_norm": 6.770204900870094, |
| "learning_rate": 8.560259645467927e-07, |
| "loss": 0.3994, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.9747530186608123, |
| "grad_norm": 6.36111561811909, |
| "learning_rate": 8.551288693101808e-07, |
| "loss": 0.3578, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.9769484083424808, |
| "grad_norm": 5.460698588059828, |
| "learning_rate": 8.542294611068573e-07, |
| "loss": 0.357, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.9791437980241493, |
| "grad_norm": 7.059061442819567, |
| "learning_rate": 8.533277457946988e-07, |
| "loss": 0.3943, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.9813391877058177, |
| "grad_norm": 5.455920326056064, |
| "learning_rate": 8.524237292466092e-07, |
| "loss": 0.3498, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.9835345773874863, |
| "grad_norm": 8.145756413409373, |
| "learning_rate": 8.515174173504795e-07, |
| "loss": 0.4005, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.9857299670691547, |
| "grad_norm": 6.554023882646871, |
| "learning_rate": 8.506088160091506e-07, |
| "loss": 0.4014, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.9879253567508233, |
| "grad_norm": 7.592409507167033, |
| "learning_rate": 8.49697931140375e-07, |
| "loss": 0.3976, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.9879253567508233, |
| "eval_accuracy": 0.776, |
| "eval_loss": 0.38674676418304443, |
| "eval_runtime": 52.2498, |
| "eval_samples_per_second": 9.569, |
| "eval_steps_per_second": 1.206, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.9901207464324918, |
| "grad_norm": 6.445935731601792, |
| "learning_rate": 8.487847686767771e-07, |
| "loss": 0.3376, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.9923161361141603, |
| "grad_norm": 7.661577003586348, |
| "learning_rate": 8.478693345658165e-07, |
| "loss": 0.3673, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.9945115257958288, |
| "grad_norm": 6.07812754087199, |
| "learning_rate": 8.469516347697472e-07, |
| "loss": 0.3901, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.9967069154774972, |
| "grad_norm": 6.985188252913582, |
| "learning_rate": 8.460316752655798e-07, |
| "loss": 0.3532, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.9989023051591658, |
| "grad_norm": 6.960412464204759, |
| "learning_rate": 8.451094620450431e-07, |
| "loss": 0.3584, |
| "step": 455 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 6.960412464204759, |
| "learning_rate": 8.441850011145435e-07, |
| "loss": 0.3955, |
| "step": 456 |
| }, |
| { |
| "epoch": 1.0021953896816684, |
| "grad_norm": 10.21658360278486, |
| "learning_rate": 8.432582984951276e-07, |
| "loss": 0.3445, |
| "step": 457 |
| }, |
| { |
| "epoch": 1.004390779363337, |
| "grad_norm": 5.546577685721332, |
| "learning_rate": 8.423293602224417e-07, |
| "loss": 0.4039, |
| "step": 458 |
| }, |
| { |
| "epoch": 1.0065861690450055, |
| "grad_norm": 5.367108573957618, |
| "learning_rate": 8.413981923466932e-07, |
| "loss": 0.4015, |
| "step": 459 |
| }, |
| { |
| "epoch": 1.008781558726674, |
| "grad_norm": 7.677188675606839, |
| "learning_rate": 8.404648009326111e-07, |
| "loss": 0.4446, |
| "step": 460 |
| }, |
| { |
| "epoch": 1.0109769484083424, |
| "grad_norm": 5.110493371628498, |
| "learning_rate": 8.395291920594061e-07, |
| "loss": 0.3263, |
| "step": 461 |
| }, |
| { |
| "epoch": 1.013172338090011, |
| "grad_norm": 5.952324335313491, |
| "learning_rate": 8.385913718207313e-07, |
| "loss": 0.3865, |
| "step": 462 |
| }, |
| { |
| "epoch": 1.0153677277716795, |
| "grad_norm": 6.4399342076113495, |
| "learning_rate": 8.376513463246429e-07, |
| "loss": 0.3821, |
| "step": 463 |
| }, |
| { |
| "epoch": 1.017563117453348, |
| "grad_norm": 5.698634912941117, |
| "learning_rate": 8.367091216935596e-07, |
| "loss": 0.4065, |
| "step": 464 |
| }, |
| { |
| "epoch": 1.0197585071350164, |
| "grad_norm": 6.689636508179602, |
| "learning_rate": 8.357647040642231e-07, |
| "loss": 0.3466, |
| "step": 465 |
| }, |
| { |
| "epoch": 1.021953896816685, |
| "grad_norm": 5.443917132674302, |
| "learning_rate": 8.348180995876587e-07, |
| "loss": 0.3785, |
| "step": 466 |
| }, |
| { |
| "epoch": 1.0241492864983535, |
| "grad_norm": 5.443111782189951, |
| "learning_rate": 8.338693144291342e-07, |
| "loss": 0.3985, |
| "step": 467 |
| }, |
| { |
| "epoch": 1.026344676180022, |
| "grad_norm": 5.887332270970953, |
| "learning_rate": 8.329183547681205e-07, |
| "loss": 0.3742, |
| "step": 468 |
| }, |
| { |
| "epoch": 1.0285400658616906, |
| "grad_norm": 6.187509225148951, |
| "learning_rate": 8.319652267982508e-07, |
| "loss": 0.3716, |
| "step": 469 |
| }, |
| { |
| "epoch": 1.030735455543359, |
| "grad_norm": 6.121985990165869, |
| "learning_rate": 8.310099367272812e-07, |
| "loss": 0.429, |
| "step": 470 |
| }, |
| { |
| "epoch": 1.0329308452250274, |
| "grad_norm": 11.523110274065871, |
| "learning_rate": 8.30052490777049e-07, |
| "loss": 0.4404, |
| "step": 471 |
| }, |
| { |
| "epoch": 1.0351262349066959, |
| "grad_norm": 5.907917517927115, |
| "learning_rate": 8.29092895183433e-07, |
| "loss": 0.3681, |
| "step": 472 |
| }, |
| { |
| "epoch": 1.0373216245883645, |
| "grad_norm": 5.901926465788518, |
| "learning_rate": 8.281311561963129e-07, |
| "loss": 0.3975, |
| "step": 473 |
| }, |
| { |
| "epoch": 1.039517014270033, |
| "grad_norm": 5.906068428010956, |
| "learning_rate": 8.271672800795284e-07, |
| "loss": 0.3665, |
| "step": 474 |
| }, |
| { |
| "epoch": 1.0417124039517014, |
| "grad_norm": 5.294577975700778, |
| "learning_rate": 8.26201273110838e-07, |
| "loss": 0.4487, |
| "step": 475 |
| }, |
| { |
| "epoch": 1.0439077936333698, |
| "grad_norm": 5.978237084104336, |
| "learning_rate": 8.252331415818788e-07, |
| "loss": 0.333, |
| "step": 476 |
| }, |
| { |
| "epoch": 1.0461031833150385, |
| "grad_norm": 5.498099104199104, |
| "learning_rate": 8.242628917981253e-07, |
| "loss": 0.3685, |
| "step": 477 |
| }, |
| { |
| "epoch": 1.048298572996707, |
| "grad_norm": 6.789539587285106, |
| "learning_rate": 8.232905300788484e-07, |
| "loss": 0.3657, |
| "step": 478 |
| }, |
| { |
| "epoch": 1.0504939626783754, |
| "grad_norm": 7.957958255256437, |
| "learning_rate": 8.223160627570736e-07, |
| "loss": 0.3326, |
| "step": 479 |
| }, |
| { |
| "epoch": 1.0526893523600438, |
| "grad_norm": 5.600567494048053, |
| "learning_rate": 8.213394961795406e-07, |
| "loss": 0.3681, |
| "step": 480 |
| }, |
| { |
| "epoch": 1.0548847420417125, |
| "grad_norm": 7.539054348927877, |
| "learning_rate": 8.203608367066615e-07, |
| "loss": 0.4324, |
| "step": 481 |
| }, |
| { |
| "epoch": 1.057080131723381, |
| "grad_norm": 8.903288366040119, |
| "learning_rate": 8.193800907124798e-07, |
| "loss": 0.4113, |
| "step": 482 |
| }, |
| { |
| "epoch": 1.0592755214050493, |
| "grad_norm": 8.111853302663592, |
| "learning_rate": 8.183972645846282e-07, |
| "loss": 0.4387, |
| "step": 483 |
| }, |
| { |
| "epoch": 1.061470911086718, |
| "grad_norm": 9.450203257580068, |
| "learning_rate": 8.174123647242877e-07, |
| "loss": 0.3703, |
| "step": 484 |
| }, |
| { |
| "epoch": 1.0636663007683864, |
| "grad_norm": 7.7233909149018025, |
| "learning_rate": 8.164253975461453e-07, |
| "loss": 0.3967, |
| "step": 485 |
| }, |
| { |
| "epoch": 1.0658616904500549, |
| "grad_norm": 6.1052811281532176, |
| "learning_rate": 8.154363694783526e-07, |
| "loss": 0.3528, |
| "step": 486 |
| }, |
| { |
| "epoch": 1.0680570801317233, |
| "grad_norm": 6.985081806859129, |
| "learning_rate": 8.14445286962484e-07, |
| "loss": 0.3402, |
| "step": 487 |
| }, |
| { |
| "epoch": 1.070252469813392, |
| "grad_norm": 6.737047071565736, |
| "learning_rate": 8.134521564534947e-07, |
| "loss": 0.3484, |
| "step": 488 |
| }, |
| { |
| "epoch": 1.0724478594950604, |
| "grad_norm": 8.024277201928818, |
| "learning_rate": 8.124569844196779e-07, |
| "loss": 0.4149, |
| "step": 489 |
| }, |
| { |
| "epoch": 1.0746432491767288, |
| "grad_norm": 6.684001008171824, |
| "learning_rate": 8.11459777342624e-07, |
| "loss": 0.3545, |
| "step": 490 |
| }, |
| { |
| "epoch": 1.0768386388583973, |
| "grad_norm": 8.174576979591736, |
| "learning_rate": 8.104605417171776e-07, |
| "loss": 0.4058, |
| "step": 491 |
| }, |
| { |
| "epoch": 1.079034028540066, |
| "grad_norm": 6.567306113947398, |
| "learning_rate": 8.094592840513949e-07, |
| "loss": 0.3559, |
| "step": 492 |
| }, |
| { |
| "epoch": 1.0812294182217344, |
| "grad_norm": 5.861714140854109, |
| "learning_rate": 8.084560108665023e-07, |
| "loss": 0.3705, |
| "step": 493 |
| }, |
| { |
| "epoch": 1.0834248079034028, |
| "grad_norm": 5.948783706558882, |
| "learning_rate": 8.074507286968528e-07, |
| "loss": 0.3855, |
| "step": 494 |
| }, |
| { |
| "epoch": 1.0856201975850714, |
| "grad_norm": 9.048770250869833, |
| "learning_rate": 8.064434440898844e-07, |
| "loss": 0.3838, |
| "step": 495 |
| }, |
| { |
| "epoch": 1.0878155872667399, |
| "grad_norm": 6.4254457902019375, |
| "learning_rate": 8.054341636060766e-07, |
| "loss": 0.3616, |
| "step": 496 |
| }, |
| { |
| "epoch": 1.0900109769484083, |
| "grad_norm": 8.363722938696041, |
| "learning_rate": 8.044228938189088e-07, |
| "loss": 0.4117, |
| "step": 497 |
| }, |
| { |
| "epoch": 1.0922063666300768, |
| "grad_norm": 8.80748939543165, |
| "learning_rate": 8.034096413148161e-07, |
| "loss": 0.3498, |
| "step": 498 |
| }, |
| { |
| "epoch": 1.0944017563117454, |
| "grad_norm": 6.142541366225709, |
| "learning_rate": 8.023944126931475e-07, |
| "loss": 0.3959, |
| "step": 499 |
| }, |
| { |
| "epoch": 1.0965971459934138, |
| "grad_norm": 6.217108697055721, |
| "learning_rate": 8.013772145661224e-07, |
| "loss": 0.4058, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0965971459934138, |
| "eval_accuracy": 0.79, |
| "eval_loss": 0.3937423825263977, |
| "eval_runtime": 52.5967, |
| "eval_samples_per_second": 9.506, |
| "eval_steps_per_second": 1.198, |
| "step": 500 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 1368, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
