| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9786476868327402, | |
| "eval_steps": 500, | |
| "global_step": 3300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0008896797153024911, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 6e-05, | |
| "loss": 1.0103, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0017793594306049821, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 0.00012, | |
| "loss": 0.9297, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0026690391459074734, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00018, | |
| "loss": 0.9383, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0035587188612099642, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00019999982536383071, | |
| "loss": 0.9338, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.004448398576512456, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00019999890852560968, | |
| "loss": 0.8937, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.005338078291814947, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00019999720583349016, | |
| "loss": 0.8876, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.006227758007117438, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.000199994717300853, | |
| "loss": 0.8804, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0071174377224199285, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00019999144294725462, | |
| "loss": 0.8889, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.00800711743772242, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.0001999873827984269, | |
| "loss": 0.8664, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.008896797153024912, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00019998253688627705, | |
| "loss": 0.876, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.009786476868327402, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00019997690524888734, | |
| "loss": 0.8375, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.010676156583629894, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.0001999704879305146, | |
| "loss": 0.8698, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.011565836298932384, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001999632849815902, | |
| "loss": 0.8346, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.012455516014234875, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00019995529645871934, | |
| "loss": 0.8573, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.013345195729537367, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.0001999465224246809, | |
| "loss": 0.8411, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.014234875444839857, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00019993696294842668, | |
| "loss": 0.8484, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.015124555160142349, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.000199926618105081, | |
| "loss": 0.8172, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.01601423487544484, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00019991548797594015, | |
| "loss": 0.8481, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.016903914590747332, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.0001999035726484716, | |
| "loss": 0.8458, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.017793594306049824, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00019989087221631343, | |
| "loss": 0.8321, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.018683274021352312, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00019987738677927365, | |
| "loss": 0.8511, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.019572953736654804, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00019986311644332915, | |
| "loss": 0.8324, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.020462633451957295, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00019984806132062517, | |
| "loss": 0.8368, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.021352313167259787, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00019983222152947428, | |
| "loss": 0.8062, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.02224199288256228, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.0001998155971943555, | |
| "loss": 0.8329, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.023131672597864767, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00019979818844591317, | |
| "loss": 0.8214, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.02402135231316726, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00019977999542095617, | |
| "loss": 0.8077, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.02491103202846975, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.0001997610182624566, | |
| "loss": 0.8004, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.025800711743772242, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.0001997412571195489, | |
| "loss": 0.8649, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.026690391459074734, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.0001997207121475284, | |
| "loss": 0.82, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.027580071174377226, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00019969938350785035, | |
| "loss": 0.8353, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.028469750889679714, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00019967727136812856, | |
| "loss": 0.8006, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.029359430604982206, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.000199654375902134, | |
| "loss": 0.8127, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.030249110320284697, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019963069728979357, | |
| "loss": 0.8029, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.03113879003558719, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00019960623571718862, | |
| "loss": 0.8162, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.03202846975088968, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001995809913765534, | |
| "loss": 0.7928, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.03291814946619217, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00019955496446627375, | |
| "loss": 0.8011, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.033807829181494664, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001995281551908854, | |
| "loss": 0.7874, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.03469750889679715, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019950056376107238, | |
| "loss": 0.827, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.03558718861209965, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00019947219039366537, | |
| "loss": 0.7829, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.036476868327402136, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019944303531164005, | |
| "loss": 0.8184, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.037366548042704624, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019941309874411524, | |
| "loss": 0.8005, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.03825622775800712, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001993823809263512, | |
| "loss": 0.8199, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.03914590747330961, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00019935088209974773, | |
| "loss": 0.8066, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.0400355871886121, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.0001993186025118423, | |
| "loss": 0.7823, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.04092526690391459, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.00019928554241630802, | |
| "loss": 0.7799, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.04181494661921708, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.0001992517020729519, | |
| "loss": 0.7996, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.042704626334519574, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019921708174771237, | |
| "loss": 0.8068, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.04359430604982206, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00019918168171265764, | |
| "loss": 0.8197, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.04448398576512456, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.0001991455022459833, | |
| "loss": 0.8255, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.045373665480427046, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019910854363201018, | |
| "loss": 0.823, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.046263345195729534, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00019907080616118222, | |
| "loss": 0.7878, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.04715302491103203, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00019903229013006394, | |
| "loss": 0.7548, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.04804270462633452, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00019899299584133845, | |
| "loss": 0.7965, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.04893238434163701, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001989529236038048, | |
| "loss": 0.7848, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0498220640569395, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001989120737323757, | |
| "loss": 0.7979, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.050711743772241996, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00019887044654807488, | |
| "loss": 0.747, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.051601423487544484, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00019882804237803488, | |
| "loss": 0.811, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.05249110320284697, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00019878486155549405, | |
| "loss": 0.7585, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.05338078291814947, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.0001987409044197943, | |
| "loss": 0.7964, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.054270462633451956, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.0001986961713163783, | |
| "loss": 0.7923, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.05516014234875445, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.0001986506625967867, | |
| "loss": 0.7821, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.05604982206405694, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00019860437861865546, | |
| "loss": 0.75, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.05693950177935943, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00019855731974571298, | |
| "loss": 0.7976, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.05782918149466192, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.0001985094863477773, | |
| "loss": 0.7894, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.05871886120996441, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00019846087880075314, | |
| "loss": 0.7952, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.059608540925266906, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00019841149748662894, | |
| "loss": 0.8058, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.060498220640569395, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.000198361342793474, | |
| "loss": 0.7727, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.06138790035587188, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00019831041511543515, | |
| "loss": 0.7792, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.06227758007117438, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00019825871485273396, | |
| "loss": 0.8123, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.06316725978647687, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00019820624241166334, | |
| "loss": 0.7966, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.06405693950177936, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00019815299820458458, | |
| "loss": 0.7883, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.06494661921708185, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019809898264992385, | |
| "loss": 0.7707, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.06583629893238434, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.0001980441961721692, | |
| "loss": 0.788, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.06672597864768683, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00019798863920186696, | |
| "loss": 0.7759, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.06761565836298933, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001979323121756185, | |
| "loss": 0.7782, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.06850533807829182, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.0001978752155360768, | |
| "loss": 0.7417, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.0693950177935943, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00019781734973194293, | |
| "loss": 0.7678, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.07028469750889679, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00019775871521796252, | |
| "loss": 0.7867, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.0711743772241993, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00019769931245492222, | |
| "loss": 0.807, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.07206405693950178, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00019763914190964609, | |
| "loss": 0.7683, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.07295373665480427, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.0001975782040549918, | |
| "loss": 0.7412, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.07384341637010676, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.0001975164993698471, | |
| "loss": 0.7793, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.07473309608540925, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019745402833912598, | |
| "loss": 0.8096, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.07562277580071175, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00019739079145376484, | |
| "loss": 0.771, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.07651245551601424, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.0001973267892107186, | |
| "loss": 0.776, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.07740213523131673, | |
| "grad_norm": 0.76171875, | |
| "learning_rate": 0.00019726202211295686, | |
| "loss": 0.7794, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.07829181494661921, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 0.00019719649066945996, | |
| "loss": 0.7751, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.0791814946619217, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.000197130195395215, | |
| "loss": 0.7957, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.0800711743772242, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.0001970631368112115, | |
| "loss": 0.7747, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.0809608540925267, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00019699531544443784, | |
| "loss": 0.7677, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.08185053380782918, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019692673182787666, | |
| "loss": 0.7749, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.08274021352313167, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00019685738650050086, | |
| "loss": 0.7795, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.08362989323843416, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00019678728000726935, | |
| "loss": 0.7748, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.08451957295373666, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001967164128991227, | |
| "loss": 0.7842, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.08540925266903915, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.000196644785732979, | |
| "loss": 0.7675, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.08629893238434164, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019657239907172925, | |
| "loss": 0.7593, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.08718861209964412, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.00019649925348423305, | |
| "loss": 0.7741, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.08807829181494661, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.0001964253495453141, | |
| "loss": 0.7842, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.08896797153024912, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00019635068783575578, | |
| "loss": 0.7615, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.0898576512455516, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001962752689422964, | |
| "loss": 0.7853, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.09074733096085409, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00019619909345762476, | |
| "loss": 0.7412, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.09163701067615658, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00019612216198037542, | |
| "loss": 0.7758, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.09252669039145907, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00019604447511512396, | |
| "loss": 0.7994, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.09341637010676157, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.00019596603347238234, | |
| "loss": 0.746, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.09430604982206406, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00019588683766859398, | |
| "loss": 0.7644, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.09519572953736655, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.000195806888326129, | |
| "loss": 0.7574, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.09608540925266904, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00019572618607327925, | |
| "loss": 0.7797, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.09697508896797152, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019564473154425348, | |
| "loss": 0.7519, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.09786476868327403, | |
| "grad_norm": 26.375, | |
| "learning_rate": 0.00019556252537917225, | |
| "loss": 0.8056, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.09875444839857651, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.000195479568224063, | |
| "loss": 0.7589, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.099644128113879, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00019539586073085482, | |
| "loss": 0.7666, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.10053380782918149, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019531140355737354, | |
| "loss": 0.7903, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.10142348754448399, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019522619736733637, | |
| "loss": 0.7793, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.10231316725978648, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00019514024283034682, | |
| "loss": 0.7505, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.10320284697508897, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00019505354062188931, | |
| "loss": 0.7642, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.10409252669039146, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00019496609142332397, | |
| "loss": 0.7675, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.10498220640569395, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00019487789592188124, | |
| "loss": 0.7635, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.10587188612099645, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00019478895481065645, | |
| "loss": 0.7739, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.10676156583629894, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019469926878860444, | |
| "loss": 0.7302, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.10765124555160142, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.000194608838560534, | |
| "loss": 0.7563, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.10854092526690391, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00019451766483710237, | |
| "loss": 0.7473, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.1094306049822064, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00019442574833480962, | |
| "loss": 0.7352, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.1103202846975089, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00019433308977599305, | |
| "loss": 0.7561, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.11120996441281139, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001942396898888215, | |
| "loss": 0.7633, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.11209964412811388, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00019414554940728963, | |
| "loss": 0.7221, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.11298932384341637, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001940506690712122, | |
| "loss": 0.7552, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.11387900355871886, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001939550496262181, | |
| "loss": 0.7563, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.11476868327402136, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00019385869182374474, | |
| "loss": 0.7643, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.11565836298932385, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001937615964210319, | |
| "loss": 0.7562, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.11654804270462633, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00019366376418111588, | |
| "loss": 0.7425, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.11743772241992882, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001935651958728236, | |
| "loss": 0.7514, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.11832740213523131, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001934658922707664, | |
| "loss": 0.7466, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.11921708185053381, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00019336585415533398, | |
| "loss": 0.7557, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.1201067615658363, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00019326508231268838, | |
| "loss": 0.7685, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.12099644128113879, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00019316357753475772, | |
| "loss": 0.7632, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.12188612099644128, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00019306134061922994, | |
| "loss": 0.7484, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.12277580071174377, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001929583723695466, | |
| "loss": 0.7623, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.12366548042704627, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00019285467359489653, | |
| "loss": 0.7662, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.12455516014234876, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001927502451102095, | |
| "loss": 0.7346, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.12544483985765126, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00019264508773614972, | |
| "loss": 0.7275, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.12633451957295375, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019253920229910965, | |
| "loss": 0.7476, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.12722419928825623, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00019243258963120313, | |
| "loss": 0.7471, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.12811387900355872, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00019232525057025915, | |
| "loss": 0.7575, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.1290035587188612, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00019221718595981507, | |
| "loss": 0.7385, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.1298932384341637, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00019210839664911012, | |
| "loss": 0.7562, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.1307829181494662, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019199888349307872, | |
| "loss": 0.7409, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.13167259786476868, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001918886473523436, | |
| "loss": 0.7473, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.13256227758007116, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00019177768909320927, | |
| "loss": 0.7526, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.13345195729537365, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00019166600958765506, | |
| "loss": 0.7723, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.13434163701067617, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00019155360971332826, | |
| "loss": 0.7301, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.13523131672597866, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.0001914404903535373, | |
| "loss": 0.7629, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.13612099644128114, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00019132665239724486, | |
| "loss": 0.7431, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.13701067615658363, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00019121209673906065, | |
| "loss": 0.742, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.13790035587188612, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.0001910968242792346, | |
| "loss": 0.7517, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.1387900355871886, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00019098083592364974, | |
| "loss": 0.7371, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.1396797153024911, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.00019086413258381506, | |
| "loss": 0.7498, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.14056939501779359, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00019074671517685827, | |
| "loss": 0.7289, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.14145907473309607, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019062858462551874, | |
| "loss": 0.7158, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.1423487544483986, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 0.0001905097418581401, | |
| "loss": 0.729, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.14323843416370108, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00019039018780866312, | |
| "loss": 0.7763, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.14412811387900357, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00019026992341661817, | |
| "loss": 0.7581, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.14501779359430605, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 0.00019014894962711805, | |
| "loss": 0.7664, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.14590747330960854, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.00019002726739085027, | |
| "loss": 0.7466, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.14679715302491103, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 0.00018990487766406994, | |
| "loss": 0.7463, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.14768683274021352, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.000189781781408592, | |
| "loss": 0.7512, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.148576512455516, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001896579795917837, | |
| "loss": 0.7208, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.1494661921708185, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00018953347318655705, | |
| "loss": 0.7514, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.15035587188612098, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001894082631713612, | |
| "loss": 0.7525, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.1512455516014235, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00018928235053017472, | |
| "loss": 0.7335, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.152135231316726, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001891557362524977, | |
| "loss": 0.7331, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.15302491103202848, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018902842133334432, | |
| "loss": 0.7437, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.15391459074733096, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00018890040677323474, | |
| "loss": 0.7029, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.15480427046263345, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.0001887716935781873, | |
| "loss": 0.7364, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.15569395017793594, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00018864228275971076, | |
| "loss": 0.7084, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.15658362989323843, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018851217533479616, | |
| "loss": 0.7301, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.15747330960854092, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00018838137232590895, | |
| "loss": 0.7352, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.1583629893238434, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00018824987476098092, | |
| "loss": 0.7245, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.1592526690391459, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00018811768367340198, | |
| "loss": 0.7509, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.1601423487544484, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.0001879848001020124, | |
| "loss": 0.732, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.1610320284697509, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00018785122509109426, | |
| "loss": 0.7194, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.1619217081850534, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00018771695969036344, | |
| "loss": 0.7389, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.16281138790035588, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00018758200495496132, | |
| "loss": 0.7155, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.16370106761565836, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0001874463619454466, | |
| "loss": 0.7561, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.16459074733096085, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0001873100317277867, | |
| "loss": 0.7208, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.16548042704626334, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00018717301537334973, | |
| "loss": 0.682, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.16637010676156583, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00018703531395889575, | |
| "loss": 0.7293, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.16725978647686832, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018689692856656853, | |
| "loss": 0.7452, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.16814946619217083, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00018675786028388692, | |
| "loss": 0.73, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.16903914590747332, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0001866181102037364, | |
| "loss": 0.728, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.1699288256227758, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018647767942436038, | |
| "loss": 0.7141, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.1708185053380783, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001863365690493517, | |
| "loss": 0.728, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.17170818505338079, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00018619478018764378, | |
| "loss": 0.7351, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.17259786476868327, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00018605231395350214, | |
| "loss": 0.7291, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.17348754448398576, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00018590917146651544, | |
| "loss": 0.7626, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.17437722419928825, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00018576535385158674, | |
| "loss": 0.7216, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.17526690391459074, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00018562086223892474, | |
| "loss": 0.7437, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.17615658362989323, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.0001854756977640348, | |
| "loss": 0.7354, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.17704626334519574, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018532986156771008, | |
| "loss": 0.7039, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.17793594306049823, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00018518335479602248, | |
| "loss": 0.7506, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.17882562277580072, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00018503617860031376, | |
| "loss": 0.748, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.1797153024911032, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00018488833413718645, | |
| "loss": 0.7402, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.1806049822064057, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00018473982256849466, | |
| "loss": 0.7068, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.18149466192170818, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001845906450613351, | |
| "loss": 0.7265, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.18238434163701067, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.0001844408027880378, | |
| "loss": 0.7285, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.18327402135231316, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018429029692615701, | |
| "loss": 0.7226, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.18416370106761565, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00018413912865846178, | |
| "loss": 0.7204, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.18505338078291814, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00018398729917292684, | |
| "loss": 0.724, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.18594306049822065, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00018383480966272306, | |
| "loss": 0.7032, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.18683274021352314, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00018368166132620836, | |
| "loss": 0.7126, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.18772241992882563, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00018352785536691804, | |
| "loss": 0.7188, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.18861209964412812, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00018337339299355542, | |
| "loss": 0.7207, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.1895017793594306, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00018321827541998228, | |
| "loss": 0.7095, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.1903914590747331, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001830625038652095, | |
| "loss": 0.733, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.19128113879003558, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.0001829060795533872, | |
| "loss": 0.7192, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.19217081850533807, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018274900371379542, | |
| "loss": 0.7178, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.19306049822064056, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00018259127758083417, | |
| "loss": 0.6983, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.19395017793594305, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00018243290239401404, | |
| "loss": 0.7008, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.19483985765124556, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001822738793979461, | |
| "loss": 0.6886, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.19572953736654805, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0001821142098423325, | |
| "loss": 0.6909, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.19661921708185054, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00018195389498195627, | |
| "loss": 0.7114, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.19750889679715303, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00018179293607667178, | |
| "loss": 0.7185, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.19839857651245552, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018163133439139467, | |
| "loss": 0.7284, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.199288256227758, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018146909119609196, | |
| "loss": 0.702, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.2001779359430605, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00018130620776577198, | |
| "loss": 0.7232, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.20106761565836298, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00018114268538047456, | |
| "loss": 0.7201, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.20195729537366547, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00018097852532526074, | |
| "loss": 0.704, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.20284697508896798, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001808137288902028, | |
| "loss": 0.7134, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.20373665480427047, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001806482973703741, | |
| "loss": 0.7004, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.20462633451957296, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00018048223206583878, | |
| "loss": 0.7124, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.20551601423487545, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00018031553428164186, | |
| "loss": 0.7271, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.20640569395017794, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001801482053277987, | |
| "loss": 0.6942, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.20729537366548043, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00017998024651928464, | |
| "loss": 0.7274, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.20818505338078291, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.000179811659176025, | |
| "loss": 0.7046, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.2090747330960854, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00017964244462288448, | |
| "loss": 0.6981, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.2099644128113879, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001794726041896567, | |
| "loss": 0.7034, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.21085409252669038, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00017930213921105392, | |
| "loss": 0.7157, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.2117437722419929, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00017913105102669642, | |
| "loss": 0.7015, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.21263345195729538, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.00017895934098110207, | |
| "loss": 0.714, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.21352313167259787, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.0001787870104236757, | |
| "loss": 0.7185, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.21441281138790036, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00017861406070869844, | |
| "loss": 0.6969, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.21530249110320285, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00017844049319531725, | |
| "loss": 0.7188, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.21619217081850534, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00017826630924753408, | |
| "loss": 0.7109, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.21708185053380782, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00017809151023419516, | |
| "loss": 0.715, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.2179715302491103, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001779160975289804, | |
| "loss": 0.7103, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.2188612099644128, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00017774007251039245, | |
| "loss": 0.7083, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.2197508896797153, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00017756343656174584, | |
| "loss": 0.733, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.2206405693950178, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00017738619107115618, | |
| "loss": 0.7088, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.2215302491103203, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00017720833743152935, | |
| "loss": 0.715, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.22241992882562278, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0001770298770405503, | |
| "loss": 0.707, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.22330960854092527, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.0001768508113006723, | |
| "loss": 0.6825, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.22419928825622776, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00017667114161910586, | |
| "loss": 0.72, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.22508896797153025, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00017649086940780748, | |
| "loss": 0.7141, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.22597864768683273, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00017630999608346886, | |
| "loss": 0.6972, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.22686832740213522, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00017612852306750566, | |
| "loss": 0.6851, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.2277580071174377, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00017594645178604611, | |
| "loss": 0.7086, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.22864768683274023, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.0001757637836699202, | |
| "loss": 0.6971, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.22953736654804271, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.000175580520154648, | |
| "loss": 0.7095, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.2304270462633452, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001753966626804288, | |
| "loss": 0.6622, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.2313167259786477, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00017521221269212943, | |
| "loss": 0.7095, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.23220640569395018, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00017502717163927315, | |
| "loss": 0.7202, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.23309608540925267, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 0.0001748415409760282, | |
| "loss": 0.7216, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.23398576512455516, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00017465532216119625, | |
| "loss": 0.6692, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.23487544483985764, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.00017446851665820116, | |
| "loss": 0.6887, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.23576512455516013, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00017428112593507723, | |
| "loss": 0.7244, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.23665480427046262, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00017409315146445784, | |
| "loss": 0.7044, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.23754448398576514, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00017390459472356383, | |
| "loss": 0.7155, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.23843416370106763, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00017371545719419186, | |
| "loss": 0.698, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.2393238434163701, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00017352574036270282, | |
| "loss": 0.6877, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.2402135231316726, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00017333544572001007, | |
| "loss": 0.7019, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.2411032028469751, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00017314457476156782, | |
| "loss": 0.711, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.24199288256227758, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00017295312898735934, | |
| "loss": 0.7147, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.24288256227758007, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00017276110990188507, | |
| "loss": 0.6981, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.24377224199288255, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.000172568519014151, | |
| "loss": 0.6907, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.24466192170818504, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00017237535783765662, | |
| "loss": 0.7123, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.24555160142348753, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00017218162789038312, | |
| "loss": 0.7002, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.24644128113879005, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00017198733069478153, | |
| "loss": 0.6812, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.24733096085409254, | |
| "grad_norm": 0.330078125, | |
| "learning_rate": 0.00017179246777776052, | |
| "loss": 0.697, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.24822064056939502, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00017159704067067468, | |
| "loss": 0.6925, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.2491103202846975, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001714010509093123, | |
| "loss": 0.7084, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00017120450003388338, | |
| "loss": 0.6717, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.2508896797153025, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001710073895890075, | |
| "loss": 0.7053, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.251779359430605, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00017080972112370167, | |
| "loss": 0.6479, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.2526690391459075, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.00017061149619136817, | |
| "loss": 0.6837, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.25355871886120995, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.0001704127163497824, | |
| "loss": 0.7102, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.25444839857651247, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.0001702133831610805, | |
| "loss": 0.7029, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.25533807829181493, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00017001349819174727, | |
| "loss": 0.6871, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.25622775800711745, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00016981306301260357, | |
| "loss": 0.6933, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.2571174377224199, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0001696120791987944, | |
| "loss": 0.7173, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.2580071174377224, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.000169410548329776, | |
| "loss": 0.6973, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.25889679715302494, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001692084719893039, | |
| "loss": 0.6909, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.2597864768683274, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001690058517654203, | |
| "loss": 0.7051, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.2606761565836299, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00016880268925044143, | |
| "loss": 0.6671, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.2615658362989324, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.0001685989860409453, | |
| "loss": 0.6904, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.2624555160142349, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00016839474373775892, | |
| "loss": 0.7076, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.26334519572953735, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00016818996394594603, | |
| "loss": 0.7155, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.26423487544483987, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00016798464827479404, | |
| "loss": 0.6727, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.26512455516014233, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001677787983378019, | |
| "loss": 0.7061, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.26601423487544484, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00016757241575266694, | |
| "loss": 0.6945, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.2669039145907473, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00016736550214127246, | |
| "loss": 0.6812, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2677935943060498, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.0001671580591296749, | |
| "loss": 0.7009, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.26868327402135234, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00016695008834809107, | |
| "loss": 0.6928, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.2695729537366548, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00016674159143088526, | |
| "loss": 0.7009, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.2704626334519573, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00016653257001655652, | |
| "loss": 0.6824, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.2713523131672598, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00016632302574772577, | |
| "loss": 0.7052, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.2722419928825623, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001661129602711227, | |
| "loss": 0.6904, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.27313167259786475, | |
| "grad_norm": 0.333984375, | |
| "learning_rate": 0.0001659023752375731, | |
| "loss": 0.6945, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.27402135231316727, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00016569127230198577, | |
| "loss": 0.6761, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.2749110320284697, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0001654796531233394, | |
| "loss": 0.7051, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.27580071174377224, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00016526751936466974, | |
| "loss": 0.6919, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.27669039145907476, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001650548726930564, | |
| "loss": 0.7075, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.2775800711743772, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00016484171477960976, | |
| "loss": 0.6621, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.27846975088967973, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0001646280472994579, | |
| "loss": 0.6956, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.2793594306049822, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00016441387193173336, | |
| "loss": 0.6791, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.2802491103202847, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00016419919035956, | |
| "loss": 0.6891, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.28113879003558717, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.0001639840042700397, | |
| "loss": 0.6948, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.2820284697508897, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00016376831535423923, | |
| "loss": 0.6941, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.28291814946619215, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.00016355212530717682, | |
| "loss": 0.6737, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.28380782918149466, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 0.00016333543582780898, | |
| "loss": 0.6966, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.2846975088967972, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00016311824861901694, | |
| "loss": 0.6938, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.28558718861209964, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00016290056538759352, | |
| "loss": 0.6963, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.28647686832740216, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00016268238784422954, | |
| "loss": 0.6931, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.2873665480427046, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00016246371770350045, | |
| "loss": 0.6698, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.28825622775800713, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00016224455668385282, | |
| "loss": 0.6909, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.2891459074733096, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001620249065075909, | |
| "loss": 0.7073, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.2900355871886121, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00016180476890086297, | |
| "loss": 0.6631, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.29092526690391457, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00016158414559364789, | |
| "loss": 0.6814, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.2918149466192171, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00016136303831974146, | |
| "loss": 0.6889, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.29270462633451955, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.0001611414488167427, | |
| "loss": 0.6689, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.29359430604982206, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00016091937882604048, | |
| "loss": 0.6845, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.2944839857651246, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00016069683009279942, | |
| "loss": 0.6819, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.29537366548042704, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.0001604738043659466, | |
| "loss": 0.6779, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.29626334519572955, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00016025030339815745, | |
| "loss": 0.6603, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.297153024911032, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00016002632894584226, | |
| "loss": 0.6968, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.29804270462633453, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00015980188276913215, | |
| "loss": 0.6903, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.298932384341637, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00015957696663186546, | |
| "loss": 0.6829, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.2998220640569395, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00015935158230157367, | |
| "loss": 0.67, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.30071174377224197, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015912573154946768, | |
| "loss": 0.7109, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.3016014234875445, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.0001588994161504238, | |
| "loss": 0.6877, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.302491103202847, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00015867263788296984, | |
| "loss": 0.6896, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.30338078291814946, | |
| "grad_norm": 0.31640625, | |
| "learning_rate": 0.00015844539852927109, | |
| "loss": 0.7024, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.304270462633452, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00015821769987511635, | |
| "loss": 0.6913, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.30516014234875444, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015798954370990393, | |
| "loss": 0.6893, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.30604982206405695, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001577609318266275, | |
| "loss": 0.6613, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.3069395017793594, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00015753186602186209, | |
| "loss": 0.6827, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.30782918149466193, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00015730234809574985, | |
| "loss": 0.6789, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.3087188612099644, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00015707237985198612, | |
| "loss": 0.6798, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.3096085409252669, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00015684196309780494, | |
| "loss": 0.6736, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.3104982206405694, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.0001566110996439652, | |
| "loss": 0.6915, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.3113879003558719, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0001563797913047361, | |
| "loss": 0.6674, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3122775800711744, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00015614803989788314, | |
| "loss": 0.6783, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.31316725978647686, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00015591584724465363, | |
| "loss": 0.6735, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.3140569395017794, | |
| "grad_norm": 0.3203125, | |
| "learning_rate": 0.00015568321516976248, | |
| "loss": 0.6884, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.31494661921708184, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00015545014550137786, | |
| "loss": 0.6765, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.31583629893238435, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00015521664007110691, | |
| "loss": 0.6778, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.3167259786476868, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00015498270071398116, | |
| "loss": 0.6766, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.31761565836298933, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00015474832926844223, | |
| "loss": 0.6813, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.3185053380782918, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015451352757632733, | |
| "loss": 0.6756, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.3193950177935943, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0001542782974828549, | |
| "loss": 0.6401, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.3202846975088968, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00015404264083660992, | |
| "loss": 0.6928, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.3211743772241993, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00015380655948952961, | |
| "loss": 0.6618, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.3220640569395018, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00015357005529688866, | |
| "loss": 0.7026, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.32295373665480426, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00015333313011728478, | |
| "loss": 0.6659, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.3238434163701068, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00015309578581262402, | |
| "loss": 0.6674, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.32473309608540923, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00015285802424810626, | |
| "loss": 0.6893, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.32562277580071175, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015261984729221038, | |
| "loss": 0.6829, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.3265124555160142, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 0.00015238125681667973, | |
| "loss": 0.6579, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.3274021352313167, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00015214225469650726, | |
| "loss": 0.6546, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.32829181494661924, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015190284280992107, | |
| "loss": 0.6665, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.3291814946619217, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00015166302303836927, | |
| "loss": 0.6877, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.3300711743772242, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00015142279726650543, | |
| "loss": 0.6823, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.3309608540925267, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.0001511821673821738, | |
| "loss": 0.6788, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.3318505338078292, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001509411352763943, | |
| "loss": 0.6804, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.33274021352313166, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00015069970284334785, | |
| "loss": 0.6902, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.33362989323843417, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00015045787198036132, | |
| "loss": 0.6713, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.33451957295373663, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001502156445878927, | |
| "loss": 0.6787, | |
| "step": 1128 | |
| }, | |
| { | |
| "epoch": 0.33540925266903915, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00014997302256951624, | |
| "loss": 0.6657, | |
| "step": 1131 | |
| }, | |
| { | |
| "epoch": 0.33629893238434166, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00014973000783190726, | |
| "loss": 0.657, | |
| "step": 1134 | |
| }, | |
| { | |
| "epoch": 0.3371886120996441, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00014948660228482745, | |
| "loss": 0.6838, | |
| "step": 1137 | |
| }, | |
| { | |
| "epoch": 0.33807829181494664, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00014924280784110963, | |
| "loss": 0.6536, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.3389679715302491, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00014899862641664288, | |
| "loss": 0.6654, | |
| "step": 1143 | |
| }, | |
| { | |
| "epoch": 0.3398576512455516, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0001487540599303574, | |
| "loss": 0.6894, | |
| "step": 1146 | |
| }, | |
| { | |
| "epoch": 0.3407473309608541, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001485091103042094, | |
| "loss": 0.6429, | |
| "step": 1149 | |
| }, | |
| { | |
| "epoch": 0.3416370106761566, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0001482637794631661, | |
| "loss": 0.6806, | |
| "step": 1152 | |
| }, | |
| { | |
| "epoch": 0.34252669039145905, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00014801806933519048, | |
| "loss": 0.67, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.34341637010676157, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0001477719818512263, | |
| "loss": 0.6683, | |
| "step": 1158 | |
| }, | |
| { | |
| "epoch": 0.34430604982206403, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00014752551894518272, | |
| "loss": 0.6468, | |
| "step": 1161 | |
| }, | |
| { | |
| "epoch": 0.34519572953736655, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00014727868255391924, | |
| "loss": 0.6664, | |
| "step": 1164 | |
| }, | |
| { | |
| "epoch": 0.34608540925266906, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00014703147461723041, | |
| "loss": 0.6905, | |
| "step": 1167 | |
| }, | |
| { | |
| "epoch": 0.3469750889679715, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00014678389707783071, | |
| "loss": 0.6777, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.34786476868327404, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00014653595188133904, | |
| "loss": 0.656, | |
| "step": 1173 | |
| }, | |
| { | |
| "epoch": 0.3487544483985765, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.0001462876409762637, | |
| "loss": 0.6772, | |
| "step": 1176 | |
| }, | |
| { | |
| "epoch": 0.349644128113879, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00014603896631398692, | |
| "loss": 0.667, | |
| "step": 1179 | |
| }, | |
| { | |
| "epoch": 0.3505338078291815, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.00014578992984874955, | |
| "loss": 0.6635, | |
| "step": 1182 | |
| }, | |
| { | |
| "epoch": 0.351423487544484, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00014554053353763575, | |
| "loss": 0.6726, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.35231316725978645, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00014529077934055752, | |
| "loss": 0.67, | |
| "step": 1188 | |
| }, | |
| { | |
| "epoch": 0.35320284697508897, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00014504066922023934, | |
| "loss": 0.6716, | |
| "step": 1191 | |
| }, | |
| { | |
| "epoch": 0.3540925266903915, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00014479020514220284, | |
| "loss": 0.6786, | |
| "step": 1194 | |
| }, | |
| { | |
| "epoch": 0.35498220640569395, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00014453938907475124, | |
| "loss": 0.6636, | |
| "step": 1197 | |
| }, | |
| { | |
| "epoch": 0.35587188612099646, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00014428822298895387, | |
| "loss": 0.6316, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.3567615658362989, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00014403670885863073, | |
| "loss": 0.6849, | |
| "step": 1203 | |
| }, | |
| { | |
| "epoch": 0.35765124555160144, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00014378484866033704, | |
| "loss": 0.6701, | |
| "step": 1206 | |
| }, | |
| { | |
| "epoch": 0.3585409252669039, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00014353264437334758, | |
| "loss": 0.629, | |
| "step": 1209 | |
| }, | |
| { | |
| "epoch": 0.3594306049822064, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00014328009797964113, | |
| "loss": 0.6799, | |
| "step": 1212 | |
| }, | |
| { | |
| "epoch": 0.3603202846975089, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00014302721146388514, | |
| "loss": 0.6566, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.3612099644128114, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00014277398681341983, | |
| "loss": 0.6607, | |
| "step": 1218 | |
| }, | |
| { | |
| "epoch": 0.3620996441281139, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001425204260182426, | |
| "loss": 0.6603, | |
| "step": 1221 | |
| }, | |
| { | |
| "epoch": 0.36298932384341637, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00014226653107099273, | |
| "loss": 0.6862, | |
| "step": 1224 | |
| }, | |
| { | |
| "epoch": 0.3638790035587189, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00014201230396693526, | |
| "loss": 0.6771, | |
| "step": 1227 | |
| }, | |
| { | |
| "epoch": 0.36476868327402134, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00014175774670394562, | |
| "loss": 0.6682, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.36565836298932386, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001415028612824938, | |
| "loss": 0.664, | |
| "step": 1233 | |
| }, | |
| { | |
| "epoch": 0.3665480427046263, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00014124764970562873, | |
| "loss": 0.6675, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 0.36743772241992884, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001409921139789624, | |
| "loss": 0.6531, | |
| "step": 1239 | |
| }, | |
| { | |
| "epoch": 0.3683274021352313, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00014073625611065423, | |
| "loss": 0.6503, | |
| "step": 1242 | |
| }, | |
| { | |
| "epoch": 0.3692170818505338, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00014048007811139513, | |
| "loss": 0.6663, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.3701067615658363, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00014022358199439192, | |
| "loss": 0.6692, | |
| "step": 1248 | |
| }, | |
| { | |
| "epoch": 0.3709964412811388, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001399667697753513, | |
| "loss": 0.636, | |
| "step": 1251 | |
| }, | |
| { | |
| "epoch": 0.3718861209964413, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00013970964347246418, | |
| "loss": 0.6724, | |
| "step": 1254 | |
| }, | |
| { | |
| "epoch": 0.37277580071174377, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001394522051063897, | |
| "loss": 0.6496, | |
| "step": 1257 | |
| }, | |
| { | |
| "epoch": 0.3736654804270463, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00013919445670023932, | |
| "loss": 0.681, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.37455516014234874, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00013893640027956106, | |
| "loss": 0.665, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 0.37544483985765126, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00013867803787232348, | |
| "loss": 0.6562, | |
| "step": 1266 | |
| }, | |
| { | |
| "epoch": 0.3763345195729537, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.0001384193715088999, | |
| "loss": 0.676, | |
| "step": 1269 | |
| }, | |
| { | |
| "epoch": 0.37722419928825623, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00013816040322205207, | |
| "loss": 0.6744, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 0.3781138790035587, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00013790113504691463, | |
| "loss": 0.661, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.3790035587188612, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00013764156902097891, | |
| "loss": 0.6443, | |
| "step": 1278 | |
| }, | |
| { | |
| "epoch": 0.3798932384341637, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00013738170718407687, | |
| "loss": 0.6613, | |
| "step": 1281 | |
| }, | |
| { | |
| "epoch": 0.3807829181494662, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001371215515783652, | |
| "loss": 0.6629, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 0.3816725978647687, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00013686110424830923, | |
| "loss": 0.6735, | |
| "step": 1287 | |
| }, | |
| { | |
| "epoch": 0.38256227758007116, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 0.00013660036724066668, | |
| "loss": 0.6597, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.3834519572953737, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00013633934260447192, | |
| "loss": 0.6712, | |
| "step": 1293 | |
| }, | |
| { | |
| "epoch": 0.38434163701067614, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00013607803239101964, | |
| "loss": 0.6494, | |
| "step": 1296 | |
| }, | |
| { | |
| "epoch": 0.38523131672597866, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00013581643865384873, | |
| "loss": 0.6474, | |
| "step": 1299 | |
| }, | |
| { | |
| "epoch": 0.3861209964412811, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0001355545634487262, | |
| "loss": 0.6386, | |
| "step": 1302 | |
| }, | |
| { | |
| "epoch": 0.38701067615658363, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.000135292408833631, | |
| "loss": 0.6541, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.3879003558718861, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00013502997686873797, | |
| "loss": 0.6444, | |
| "step": 1308 | |
| }, | |
| { | |
| "epoch": 0.3887900355871886, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00013476726961640133, | |
| "loss": 0.6295, | |
| "step": 1311 | |
| }, | |
| { | |
| "epoch": 0.3896797153024911, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.0001345042891411389, | |
| "loss": 0.6551, | |
| "step": 1314 | |
| }, | |
| { | |
| "epoch": 0.3905693950177936, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.0001342410375096155, | |
| "loss": 0.6757, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 0.3914590747330961, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00013397751679062692, | |
| "loss": 0.6516, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.39234875444839856, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00013371372905508362, | |
| "loss": 0.6425, | |
| "step": 1323 | |
| }, | |
| { | |
| "epoch": 0.3932384341637011, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00013344967637599444, | |
| "loss": 0.642, | |
| "step": 1326 | |
| }, | |
| { | |
| "epoch": 0.39412811387900354, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00013318536082845026, | |
| "loss": 0.6309, | |
| "step": 1329 | |
| }, | |
| { | |
| "epoch": 0.39501779359430605, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001329207844896078, | |
| "loss": 0.6477, | |
| "step": 1332 | |
| }, | |
| { | |
| "epoch": 0.3959074733096085, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00013265594943867327, | |
| "loss": 0.6476, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.39679715302491103, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00013239085775688592, | |
| "loss": 0.646, | |
| "step": 1338 | |
| }, | |
| { | |
| "epoch": 0.39768683274021355, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 0.00013212551152750178, | |
| "loss": 0.6614, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 0.398576512455516, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00013185991283577738, | |
| "loss": 0.6477, | |
| "step": 1344 | |
| }, | |
| { | |
| "epoch": 0.3994661921708185, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00013159406376895313, | |
| "loss": 0.6417, | |
| "step": 1347 | |
| }, | |
| { | |
| "epoch": 0.400355871886121, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00013132796641623703, | |
| "loss": 0.6358, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.4012455516014235, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00013106162286878842, | |
| "loss": 0.652, | |
| "step": 1353 | |
| }, | |
| { | |
| "epoch": 0.40213523131672596, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00013079503521970127, | |
| "loss": 0.6511, | |
| "step": 1356 | |
| }, | |
| { | |
| "epoch": 0.4030249110320285, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00013052820556398785, | |
| "loss": 0.6546, | |
| "step": 1359 | |
| }, | |
| { | |
| "epoch": 0.40391459074733094, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 0.0001302611359985623, | |
| "loss": 0.6564, | |
| "step": 1362 | |
| }, | |
| { | |
| "epoch": 0.40480427046263345, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00012999382862222415, | |
| "loss": 0.6609, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.40569395017793597, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00012972628553564177, | |
| "loss": 0.6508, | |
| "step": 1368 | |
| }, | |
| { | |
| "epoch": 0.40658362989323843, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001294585088413358, | |
| "loss": 0.652, | |
| "step": 1371 | |
| }, | |
| { | |
| "epoch": 0.40747330960854095, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00012919050064366295, | |
| "loss": 0.6492, | |
| "step": 1374 | |
| }, | |
| { | |
| "epoch": 0.4083629893238434, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00012892226304879893, | |
| "loss": 0.6136, | |
| "step": 1377 | |
| }, | |
| { | |
| "epoch": 0.4092526690391459, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00012865379816472242, | |
| "loss": 0.6543, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.4101423487544484, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.0001283851081011982, | |
| "loss": 0.6568, | |
| "step": 1383 | |
| }, | |
| { | |
| "epoch": 0.4110320284697509, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00012811619496976066, | |
| "loss": 0.6457, | |
| "step": 1386 | |
| }, | |
| { | |
| "epoch": 0.41192170818505336, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00012784706088369714, | |
| "loss": 0.6653, | |
| "step": 1389 | |
| }, | |
| { | |
| "epoch": 0.4128113879003559, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001275777079580315, | |
| "loss": 0.6545, | |
| "step": 1392 | |
| }, | |
| { | |
| "epoch": 0.41370106761565834, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00012730813830950732, | |
| "loss": 0.6465, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.41459074733096085, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00012703835405657122, | |
| "loss": 0.6304, | |
| "step": 1398 | |
| }, | |
| { | |
| "epoch": 0.41548042704626337, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.0001267683573193565, | |
| "loss": 0.6628, | |
| "step": 1401 | |
| }, | |
| { | |
| "epoch": 0.41637010676156583, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001264981502196662, | |
| "loss": 0.6328, | |
| "step": 1404 | |
| }, | |
| { | |
| "epoch": 0.41725978647686834, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 0.00012622773488095643, | |
| "loss": 0.6621, | |
| "step": 1407 | |
| }, | |
| { | |
| "epoch": 0.4181494661921708, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00012595711342831996, | |
| "loss": 0.6672, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.4190391459074733, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00012568628798846924, | |
| "loss": 0.6496, | |
| "step": 1413 | |
| }, | |
| { | |
| "epoch": 0.4199288256227758, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00012541526068971973, | |
| "loss": 0.6318, | |
| "step": 1416 | |
| }, | |
| { | |
| "epoch": 0.4208185053380783, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001251440336619733, | |
| "loss": 0.6328, | |
| "step": 1419 | |
| }, | |
| { | |
| "epoch": 0.42170818505338076, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00012487260903670135, | |
| "loss": 0.6476, | |
| "step": 1422 | |
| }, | |
| { | |
| "epoch": 0.4225978647686833, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00012460098894692822, | |
| "loss": 0.6404, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.4234875444839858, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001243291755272142, | |
| "loss": 0.6592, | |
| "step": 1428 | |
| }, | |
| { | |
| "epoch": 0.42437722419928825, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.000124057170913639, | |
| "loss": 0.6489, | |
| "step": 1431 | |
| }, | |
| { | |
| "epoch": 0.42526690391459077, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 0.00012378497724378483, | |
| "loss": 0.6669, | |
| "step": 1434 | |
| }, | |
| { | |
| "epoch": 0.4261565836298932, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00012351259665671958, | |
| "loss": 0.6477, | |
| "step": 1437 | |
| }, | |
| { | |
| "epoch": 0.42704626334519574, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00012324003129298005, | |
| "loss": 0.6202, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.4279359430604982, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00012296728329455524, | |
| "loss": 0.6655, | |
| "step": 1443 | |
| }, | |
| { | |
| "epoch": 0.4288256227758007, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00012269435480486923, | |
| "loss": 0.6526, | |
| "step": 1446 | |
| }, | |
| { | |
| "epoch": 0.4297153024911032, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001224212479687646, | |
| "loss": 0.6619, | |
| "step": 1449 | |
| }, | |
| { | |
| "epoch": 0.4306049822064057, | |
| "grad_norm": 0.318359375, | |
| "learning_rate": 0.00012214796493248563, | |
| "loss": 0.6718, | |
| "step": 1452 | |
| }, | |
| { | |
| "epoch": 0.4314946619217082, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00012187450784366101, | |
| "loss": 0.6414, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.43238434163701067, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00012160087885128745, | |
| "loss": 0.6567, | |
| "step": 1458 | |
| }, | |
| { | |
| "epoch": 0.4332740213523132, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00012132708010571252, | |
| "loss": 0.6345, | |
| "step": 1461 | |
| }, | |
| { | |
| "epoch": 0.43416370106761565, | |
| "grad_norm": 0.302734375, | |
| "learning_rate": 0.00012105311375861785, | |
| "loss": 0.6611, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 0.43505338078291816, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00012077898196300208, | |
| "loss": 0.652, | |
| "step": 1467 | |
| }, | |
| { | |
| "epoch": 0.4359430604982206, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00012050468687316419, | |
| "loss": 0.6373, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.43683274021352314, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00012023023064468637, | |
| "loss": 0.6323, | |
| "step": 1473 | |
| }, | |
| { | |
| "epoch": 0.4377224199288256, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00011995561543441709, | |
| "loss": 0.6321, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 0.4386120996441281, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011968084340045425, | |
| "loss": 0.6388, | |
| "step": 1479 | |
| }, | |
| { | |
| "epoch": 0.4395017793594306, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.0001194059167021282, | |
| "loss": 0.653, | |
| "step": 1482 | |
| }, | |
| { | |
| "epoch": 0.4403914590747331, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00011913083749998464, | |
| "loss": 0.6604, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.4412811387900356, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00011885560795576783, | |
| "loss": 0.662, | |
| "step": 1488 | |
| }, | |
| { | |
| "epoch": 0.44217081850533807, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0001185802302324035, | |
| "loss": 0.6724, | |
| "step": 1491 | |
| }, | |
| { | |
| "epoch": 0.4430604982206406, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.00011830470649398182, | |
| "loss": 0.6296, | |
| "step": 1494 | |
| }, | |
| { | |
| "epoch": 0.44395017793594305, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011802903890574046, | |
| "loss": 0.6475, | |
| "step": 1497 | |
| }, | |
| { | |
| "epoch": 0.44483985765124556, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011775322963404756, | |
| "loss": 0.6665, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.445729537366548, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.0001174772808463847, | |
| "loss": 0.6361, | |
| "step": 1503 | |
| }, | |
| { | |
| "epoch": 0.44661921708185054, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00011720119471132984, | |
| "loss": 0.6302, | |
| "step": 1506 | |
| }, | |
| { | |
| "epoch": 0.447508896797153, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00011692497339854031, | |
| "loss": 0.6552, | |
| "step": 1509 | |
| }, | |
| { | |
| "epoch": 0.4483985765124555, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00011664861907873584, | |
| "loss": 0.6436, | |
| "step": 1512 | |
| }, | |
| { | |
| "epoch": 0.44928825622775803, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011637213392368118, | |
| "loss": 0.6381, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.4501779359430605, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00011609552010616954, | |
| "loss": 0.659, | |
| "step": 1518 | |
| }, | |
| { | |
| "epoch": 0.451067615658363, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00011581877980000508, | |
| "loss": 0.6412, | |
| "step": 1521 | |
| }, | |
| { | |
| "epoch": 0.45195729537366547, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00011554191517998598, | |
| "loss": 0.637, | |
| "step": 1524 | |
| }, | |
| { | |
| "epoch": 0.452846975088968, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00011526492842188745, | |
| "loss": 0.64, | |
| "step": 1527 | |
| }, | |
| { | |
| "epoch": 0.45373665480427045, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011498782170244449, | |
| "loss": 0.6525, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.45462633451957296, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00011471059719933479, | |
| "loss": 0.6248, | |
| "step": 1533 | |
| }, | |
| { | |
| "epoch": 0.4555160142348754, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00011443325709116171, | |
| "loss": 0.6334, | |
| "step": 1536 | |
| }, | |
| { | |
| "epoch": 0.45640569395017794, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00011415580355743707, | |
| "loss": 0.6381, | |
| "step": 1539 | |
| }, | |
| { | |
| "epoch": 0.45729537366548045, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00011387823877856411, | |
| "loss": 0.6367, | |
| "step": 1542 | |
| }, | |
| { | |
| "epoch": 0.4581850533807829, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00011360056493582028, | |
| "loss": 0.6443, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.45907473309608543, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011332278421134005, | |
| "loss": 0.6342, | |
| "step": 1548 | |
| }, | |
| { | |
| "epoch": 0.4599644128113879, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011304489878809803, | |
| "loss": 0.6385, | |
| "step": 1551 | |
| }, | |
| { | |
| "epoch": 0.4608540925266904, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00011276691084989134, | |
| "loss": 0.6102, | |
| "step": 1554 | |
| }, | |
| { | |
| "epoch": 0.46174377224199287, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00011248882258132299, | |
| "loss": 0.6285, | |
| "step": 1557 | |
| }, | |
| { | |
| "epoch": 0.4626334519572954, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00011221063616778425, | |
| "loss": 0.6504, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.46352313167259784, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00011193235379543778, | |
| "loss": 0.647, | |
| "step": 1563 | |
| }, | |
| { | |
| "epoch": 0.46441281138790036, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00011165397765120033, | |
| "loss": 0.645, | |
| "step": 1566 | |
| }, | |
| { | |
| "epoch": 0.4653024911032028, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 0.00011137550992272561, | |
| "loss": 0.6311, | |
| "step": 1569 | |
| }, | |
| { | |
| "epoch": 0.46619217081850534, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001110969527983869, | |
| "loss": 0.6348, | |
| "step": 1572 | |
| }, | |
| { | |
| "epoch": 0.46708185053380785, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 0.00011081830846726021, | |
| "loss": 0.6244, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.4679715302491103, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.00011053957911910671, | |
| "loss": 0.6336, | |
| "step": 1578 | |
| }, | |
| { | |
| "epoch": 0.46886120996441283, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 0.0001102607669443558, | |
| "loss": 0.6211, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 0.4697508896797153, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.00010998187413408774, | |
| "loss": 0.6337, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 0.4706405693950178, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00010970290288001644, | |
| "loss": 0.6451, | |
| "step": 1587 | |
| }, | |
| { | |
| "epoch": 0.47153024911032027, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00010942385537447236, | |
| "loss": 0.6466, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.4724199288256228, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.00010914473381038508, | |
| "loss": 0.6324, | |
| "step": 1593 | |
| }, | |
| { | |
| "epoch": 0.47330960854092524, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00010886554038126625, | |
| "loss": 0.615, | |
| "step": 1596 | |
| }, | |
| { | |
| "epoch": 0.47419928825622776, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 0.00010858627728119226, | |
| "loss": 0.6385, | |
| "step": 1599 | |
| }, | |
| { | |
| "epoch": 0.4750889679715303, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00010830694670478705, | |
| "loss": 0.6253, | |
| "step": 1602 | |
| }, | |
| { | |
| "epoch": 0.47597864768683273, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00010802755084720479, | |
| "loss": 0.6188, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.47686832740213525, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001077480919041127, | |
| "loss": 0.6179, | |
| "step": 1608 | |
| }, | |
| { | |
| "epoch": 0.4777580071174377, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00010746857207167372, | |
| "loss": 0.6427, | |
| "step": 1611 | |
| }, | |
| { | |
| "epoch": 0.4786476868327402, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00010718899354652931, | |
| "loss": 0.6096, | |
| "step": 1614 | |
| }, | |
| { | |
| "epoch": 0.4795373665480427, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00010690935852578225, | |
| "loss": 0.648, | |
| "step": 1617 | |
| }, | |
| { | |
| "epoch": 0.4804270462633452, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00010662966920697919, | |
| "loss": 0.619, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.48131672597864766, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00010634992778809357, | |
| "loss": 0.6368, | |
| "step": 1623 | |
| }, | |
| { | |
| "epoch": 0.4822064056939502, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00010607013646750818, | |
| "loss": 0.623, | |
| "step": 1626 | |
| }, | |
| { | |
| "epoch": 0.4830960854092527, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00010579029744399809, | |
| "loss": 0.6212, | |
| "step": 1629 | |
| }, | |
| { | |
| "epoch": 0.48398576512455516, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.00010551041291671311, | |
| "loss": 0.6343, | |
| "step": 1632 | |
| }, | |
| { | |
| "epoch": 0.48487544483985767, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00010523048508516075, | |
| "loss": 0.634, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.48576512455516013, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 0.00010495051614918881, | |
| "loss": 0.5907, | |
| "step": 1638 | |
| }, | |
| { | |
| "epoch": 0.48665480427046265, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00010467050830896808, | |
| "loss": 0.6122, | |
| "step": 1641 | |
| }, | |
| { | |
| "epoch": 0.4875444839857651, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 0.0001043904637649751, | |
| "loss": 0.6175, | |
| "step": 1644 | |
| }, | |
| { | |
| "epoch": 0.4884341637010676, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00010411038471797488, | |
| "loss": 0.6488, | |
| "step": 1647 | |
| }, | |
| { | |
| "epoch": 0.4893238434163701, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00010383027336900355, | |
| "loss": 0.6245, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.4902135231316726, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.00010355013191935108, | |
| "loss": 0.6203, | |
| "step": 1653 | |
| }, | |
| { | |
| "epoch": 0.49110320284697506, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 0.000103269962570544, | |
| "loss": 0.6193, | |
| "step": 1656 | |
| }, | |
| { | |
| "epoch": 0.4919928825622776, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00010298976752432812, | |
| "loss": 0.6403, | |
| "step": 1659 | |
| }, | |
| { | |
| "epoch": 0.4928825622775801, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00010270954898265112, | |
| "loss": 0.6292, | |
| "step": 1662 | |
| }, | |
| { | |
| "epoch": 0.49377224199288255, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.00010242930914764541, | |
| "loss": 0.6404, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.49466192170818507, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 0.0001021490502216107, | |
| "loss": 0.6359, | |
| "step": 1668 | |
| }, | |
| { | |
| "epoch": 0.49555160142348753, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.0001018687744069967, | |
| "loss": 0.6442, | |
| "step": 1671 | |
| }, | |
| { | |
| "epoch": 0.49644128113879005, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 0.00010158848390638587, | |
| "loss": 0.6515, | |
| "step": 1674 | |
| }, | |
| { | |
| "epoch": 0.4973309608540925, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 0.00010130818092247607, | |
| "loss": 0.6115, | |
| "step": 1677 | |
| }, | |
| { | |
| "epoch": 0.498220640569395, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 0.0001010278676580633, | |
| "loss": 0.6434, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.4991103202846975, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 0.00010074754631602428, | |
| "loss": 0.6234, | |
| "step": 1683 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.00010046721909929928, | |
| "loss": 0.6571, | |
| "step": 1686 | |
| }, | |
| { | |
| "epoch": 0.5008896797153025, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00010018688821087474, | |
| "loss": 0.6454, | |
| "step": 1689 | |
| }, | |
| { | |
| "epoch": 0.501779359430605, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.990655585376584e-05, | |
| "loss": 0.5973, | |
| "step": 1692 | |
| }, | |
| { | |
| "epoch": 0.5026690391459074, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.962622423099942e-05, | |
| "loss": 0.6094, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.50355871886121, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.934589554559653e-05, | |
| "loss": 0.642, | |
| "step": 1698 | |
| }, | |
| { | |
| "epoch": 0.5044483985765125, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.906557200055508e-05, | |
| "loss": 0.6283, | |
| "step": 1701 | |
| }, | |
| { | |
| "epoch": 0.505338078291815, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.878525579883265e-05, | |
| "loss": 0.6277, | |
| "step": 1704 | |
| }, | |
| { | |
| "epoch": 0.5062277580071174, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.850494914332908e-05, | |
| "loss": 0.627, | |
| "step": 1707 | |
| }, | |
| { | |
| "epoch": 0.5071174377224199, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 9.822465423686917e-05, | |
| "loss": 0.6435, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.5080071174377224, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.794437328218546e-05, | |
| "loss": 0.6537, | |
| "step": 1713 | |
| }, | |
| { | |
| "epoch": 0.5088967971530249, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 9.766410848190077e-05, | |
| "loss": 0.6366, | |
| "step": 1716 | |
| }, | |
| { | |
| "epoch": 0.5097864768683275, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.738386203851101e-05, | |
| "loss": 0.625, | |
| "step": 1719 | |
| }, | |
| { | |
| "epoch": 0.5106761565836299, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.710363615436776e-05, | |
| "loss": 0.6372, | |
| "step": 1722 | |
| }, | |
| { | |
| "epoch": 0.5115658362989324, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 9.682343303166117e-05, | |
| "loss": 0.6188, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.5124555160142349, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.654325487240243e-05, | |
| "loss": 0.6223, | |
| "step": 1728 | |
| }, | |
| { | |
| "epoch": 0.5133451957295374, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.626310387840648e-05, | |
| "loss": 0.6096, | |
| "step": 1731 | |
| }, | |
| { | |
| "epoch": 0.5142348754448398, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.598298225127498e-05, | |
| "loss": 0.615, | |
| "step": 1734 | |
| }, | |
| { | |
| "epoch": 0.5151245551601423, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.570289219237858e-05, | |
| "loss": 0.6377, | |
| "step": 1737 | |
| }, | |
| { | |
| "epoch": 0.5160142348754448, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.542283590284002e-05, | |
| "loss": 0.6191, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.5169039145907474, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.514281558351653e-05, | |
| "loss": 0.625, | |
| "step": 1743 | |
| }, | |
| { | |
| "epoch": 0.5177935943060499, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.486283343498277e-05, | |
| "loss": 0.6202, | |
| "step": 1746 | |
| }, | |
| { | |
| "epoch": 0.5186832740213523, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 9.458289165751339e-05, | |
| "loss": 0.613, | |
| "step": 1749 | |
| }, | |
| { | |
| "epoch": 0.5195729537366548, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.430299245106573e-05, | |
| "loss": 0.6306, | |
| "step": 1752 | |
| }, | |
| { | |
| "epoch": 0.5204626334519573, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.402313801526267e-05, | |
| "loss": 0.6162, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.5213523131672598, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 9.37433305493752e-05, | |
| "loss": 0.6472, | |
| "step": 1758 | |
| }, | |
| { | |
| "epoch": 0.5222419928825622, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.346357225230519e-05, | |
| "loss": 0.6347, | |
| "step": 1761 | |
| }, | |
| { | |
| "epoch": 0.5231316725978647, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.318386532256807e-05, | |
| "loss": 0.6155, | |
| "step": 1764 | |
| }, | |
| { | |
| "epoch": 0.5240213523131673, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.290421195827572e-05, | |
| "loss": 0.6112, | |
| "step": 1767 | |
| }, | |
| { | |
| "epoch": 0.5249110320284698, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 9.262461435711898e-05, | |
| "loss": 0.6252, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.5258007117437722, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 9.234507471635043e-05, | |
| "loss": 0.6241, | |
| "step": 1773 | |
| }, | |
| { | |
| "epoch": 0.5266903914590747, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.206559523276731e-05, | |
| "loss": 0.6346, | |
| "step": 1776 | |
| }, | |
| { | |
| "epoch": 0.5275800711743772, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 9.178617810269388e-05, | |
| "loss": 0.6489, | |
| "step": 1779 | |
| }, | |
| { | |
| "epoch": 0.5284697508896797, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.150682552196462e-05, | |
| "loss": 0.6247, | |
| "step": 1782 | |
| }, | |
| { | |
| "epoch": 0.5293594306049823, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.12275396859066e-05, | |
| "loss": 0.5991, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.5302491103202847, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.094832278932238e-05, | |
| "loss": 0.6174, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 0.5311387900355872, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.066917702647284e-05, | |
| "loss": 0.6204, | |
| "step": 1791 | |
| }, | |
| { | |
| "epoch": 0.5320284697508897, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.039010459105974e-05, | |
| "loss": 0.584, | |
| "step": 1794 | |
| }, | |
| { | |
| "epoch": 0.5329181494661922, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 9.011110767620865e-05, | |
| "loss": 0.6179, | |
| "step": 1797 | |
| }, | |
| { | |
| "epoch": 0.5338078291814946, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 8.983218847445157e-05, | |
| "loss": 0.5993, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5346975088967971, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.955334917770993e-05, | |
| "loss": 0.6324, | |
| "step": 1803 | |
| }, | |
| { | |
| "epoch": 0.5355871886120996, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 8.927459197727712e-05, | |
| "loss": 0.6048, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 0.5364768683274022, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 8.899591906380131e-05, | |
| "loss": 0.6138, | |
| "step": 1809 | |
| }, | |
| { | |
| "epoch": 0.5373665480427047, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.871733262726846e-05, | |
| "loss": 0.626, | |
| "step": 1812 | |
| }, | |
| { | |
| "epoch": 0.5382562277580071, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 8.843883485698474e-05, | |
| "loss": 0.6169, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.5391459074733096, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 8.81604279415597e-05, | |
| "loss": 0.6043, | |
| "step": 1818 | |
| }, | |
| { | |
| "epoch": 0.5400355871886121, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 8.788211406888872e-05, | |
| "loss": 0.6216, | |
| "step": 1821 | |
| }, | |
| { | |
| "epoch": 0.5409252669039146, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 8.76038954261362e-05, | |
| "loss": 0.6365, | |
| "step": 1824 | |
| }, | |
| { | |
| "epoch": 0.541814946619217, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.732577419971801e-05, | |
| "loss": 0.6226, | |
| "step": 1827 | |
| }, | |
| { | |
| "epoch": 0.5427046263345195, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 8.704775257528448e-05, | |
| "loss": 0.6179, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.5435943060498221, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 8.676983273770327e-05, | |
| "loss": 0.6191, | |
| "step": 1833 | |
| }, | |
| { | |
| "epoch": 0.5444839857651246, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 8.649201687104209e-05, | |
| "loss": 0.6243, | |
| "step": 1836 | |
| }, | |
| { | |
| "epoch": 0.5453736654804271, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.621430715855155e-05, | |
| "loss": 0.628, | |
| "step": 1839 | |
| }, | |
| { | |
| "epoch": 0.5462633451957295, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 8.593670578264814e-05, | |
| "loss": 0.6108, | |
| "step": 1842 | |
| }, | |
| { | |
| "epoch": 0.547153024911032, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 8.565921492489686e-05, | |
| "loss": 0.609, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.5480427046263345, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.538183676599426e-05, | |
| "loss": 0.6135, | |
| "step": 1848 | |
| }, | |
| { | |
| "epoch": 0.548932384341637, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.510457348575115e-05, | |
| "loss": 0.623, | |
| "step": 1851 | |
| }, | |
| { | |
| "epoch": 0.5498220640569395, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 8.482742726307569e-05, | |
| "loss": 0.6345, | |
| "step": 1854 | |
| }, | |
| { | |
| "epoch": 0.550711743772242, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.4550400275956e-05, | |
| "loss": 0.6167, | |
| "step": 1857 | |
| }, | |
| { | |
| "epoch": 0.5516014234875445, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 8.427349470144319e-05, | |
| "loss": 0.6199, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.552491103202847, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 8.399671271563438e-05, | |
| "loss": 0.589, | |
| "step": 1863 | |
| }, | |
| { | |
| "epoch": 0.5533807829181495, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 8.372005649365519e-05, | |
| "loss": 0.6183, | |
| "step": 1866 | |
| }, | |
| { | |
| "epoch": 0.5542704626334519, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 8.344352820964317e-05, | |
| "loss": 0.6555, | |
| "step": 1869 | |
| }, | |
| { | |
| "epoch": 0.5551601423487544, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.316713003673028e-05, | |
| "loss": 0.6172, | |
| "step": 1872 | |
| }, | |
| { | |
| "epoch": 0.556049822064057, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.289086414702609e-05, | |
| "loss": 0.6328, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.5569395017793595, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 8.261473271160046e-05, | |
| "loss": 0.6314, | |
| "step": 1878 | |
| }, | |
| { | |
| "epoch": 0.5578291814946619, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.233873790046684e-05, | |
| "loss": 0.6392, | |
| "step": 1881 | |
| }, | |
| { | |
| "epoch": 0.5587188612099644, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 8.206288188256486e-05, | |
| "loss": 0.6297, | |
| "step": 1884 | |
| }, | |
| { | |
| "epoch": 0.5596085409252669, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.178716682574339e-05, | |
| "loss": 0.5915, | |
| "step": 1887 | |
| }, | |
| { | |
| "epoch": 0.5604982206405694, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 8.15115948967437e-05, | |
| "loss": 0.625, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.5613879003558719, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 8.12361682611821e-05, | |
| "loss": 0.6307, | |
| "step": 1893 | |
| }, | |
| { | |
| "epoch": 0.5622775800711743, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.096088908353315e-05, | |
| "loss": 0.6018, | |
| "step": 1896 | |
| }, | |
| { | |
| "epoch": 0.5631672597864769, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 8.068575952711272e-05, | |
| "loss": 0.6045, | |
| "step": 1899 | |
| }, | |
| { | |
| "epoch": 0.5640569395017794, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.041078175406064e-05, | |
| "loss": 0.5921, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 0.5649466192170819, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.013595792532412e-05, | |
| "loss": 0.6339, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 0.5658362989323843, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 7.986129020064044e-05, | |
| "loss": 0.6147, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 0.5667259786476868, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.958678073852025e-05, | |
| "loss": 0.5981, | |
| "step": 1911 | |
| }, | |
| { | |
| "epoch": 0.5676156583629893, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 7.931243169623037e-05, | |
| "loss": 0.6201, | |
| "step": 1914 | |
| }, | |
| { | |
| "epoch": 0.5685053380782918, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.903824522977695e-05, | |
| "loss": 0.5997, | |
| "step": 1917 | |
| }, | |
| { | |
| "epoch": 0.5693950177935944, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 7.876422349388862e-05, | |
| "loss": 0.6216, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.5702846975088968, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 7.849036864199931e-05, | |
| "loss": 0.6365, | |
| "step": 1923 | |
| }, | |
| { | |
| "epoch": 0.5711743772241993, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 7.821668282623158e-05, | |
| "loss": 0.5978, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 0.5720640569395018, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 7.79431681973795e-05, | |
| "loss": 0.605, | |
| "step": 1929 | |
| }, | |
| { | |
| "epoch": 0.5729537366548043, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 7.766982690489199e-05, | |
| "loss": 0.6106, | |
| "step": 1932 | |
| }, | |
| { | |
| "epoch": 0.5738434163701067, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 7.739666109685563e-05, | |
| "loss": 0.6312, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 0.5747330960854092, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.7123672919978e-05, | |
| "loss": 0.6069, | |
| "step": 1938 | |
| }, | |
| { | |
| "epoch": 0.5756227758007118, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.685086451957084e-05, | |
| "loss": 0.6266, | |
| "step": 1941 | |
| }, | |
| { | |
| "epoch": 0.5765124555160143, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 7.657823803953288e-05, | |
| "loss": 0.5883, | |
| "step": 1944 | |
| }, | |
| { | |
| "epoch": 0.5774021352313167, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.63057956223334e-05, | |
| "loss": 0.6277, | |
| "step": 1947 | |
| }, | |
| { | |
| "epoch": 0.5782918149466192, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 7.60335394089951e-05, | |
| "loss": 0.6281, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5791814946619217, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.576147153907742e-05, | |
| "loss": 0.6225, | |
| "step": 1953 | |
| }, | |
| { | |
| "epoch": 0.5800711743772242, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.54895941506596e-05, | |
| "loss": 0.6119, | |
| "step": 1956 | |
| }, | |
| { | |
| "epoch": 0.5809608540925267, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 7.521790938032408e-05, | |
| "loss": 0.6066, | |
| "step": 1959 | |
| }, | |
| { | |
| "epoch": 0.5818505338078291, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 7.494641936313953e-05, | |
| "loss": 0.5946, | |
| "step": 1962 | |
| }, | |
| { | |
| "epoch": 0.5827402135231317, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 7.467512623264403e-05, | |
| "loss": 0.6134, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 0.5836298932384342, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 7.440403212082862e-05, | |
| "loss": 0.6004, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 0.5845195729537367, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 7.41331391581201e-05, | |
| "loss": 0.613, | |
| "step": 1971 | |
| }, | |
| { | |
| "epoch": 0.5854092526690391, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 7.386244947336462e-05, | |
| "loss": 0.6162, | |
| "step": 1974 | |
| }, | |
| { | |
| "epoch": 0.5862989323843416, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 7.359196519381092e-05, | |
| "loss": 0.6304, | |
| "step": 1977 | |
| }, | |
| { | |
| "epoch": 0.5871886120996441, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 7.33216884450934e-05, | |
| "loss": 0.6367, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.5880782918149466, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.305162135121561e-05, | |
| "loss": 0.5892, | |
| "step": 1983 | |
| }, | |
| { | |
| "epoch": 0.5889679715302492, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 7.278176603453347e-05, | |
| "loss": 0.6313, | |
| "step": 1986 | |
| }, | |
| { | |
| "epoch": 0.5898576512455516, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.251212461573873e-05, | |
| "loss": 0.6132, | |
| "step": 1989 | |
| }, | |
| { | |
| "epoch": 0.5907473309608541, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 7.224269921384206e-05, | |
| "loss": 0.6081, | |
| "step": 1992 | |
| }, | |
| { | |
| "epoch": 0.5916370106761566, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.197349194615656e-05, | |
| "loss": 0.5778, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 0.5925266903914591, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 7.170450492828125e-05, | |
| "loss": 0.6191, | |
| "step": 1998 | |
| }, | |
| { | |
| "epoch": 0.5934163701067615, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 7.143574027408408e-05, | |
| "loss": 0.6019, | |
| "step": 2001 | |
| }, | |
| { | |
| "epoch": 0.594306049822064, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 7.116720009568564e-05, | |
| "loss": 0.6255, | |
| "step": 2004 | |
| }, | |
| { | |
| "epoch": 0.5951957295373665, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.08988865034424e-05, | |
| "loss": 0.5998, | |
| "step": 2007 | |
| }, | |
| { | |
| "epoch": 0.5960854092526691, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.063080160593025e-05, | |
| "loss": 0.6059, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.5969750889679716, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.036294750992775e-05, | |
| "loss": 0.6082, | |
| "step": 2013 | |
| }, | |
| { | |
| "epoch": 0.597864768683274, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.009532632039975e-05, | |
| "loss": 0.6038, | |
| "step": 2016 | |
| }, | |
| { | |
| "epoch": 0.5987544483985765, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.982794014048077e-05, | |
| "loss": 0.6042, | |
| "step": 2019 | |
| }, | |
| { | |
| "epoch": 0.599644128113879, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.956079107145845e-05, | |
| "loss": 0.6194, | |
| "step": 2022 | |
| }, | |
| { | |
| "epoch": 0.6005338078291815, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.92938812127571e-05, | |
| "loss": 0.6125, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.6014234875444839, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.902721266192111e-05, | |
| "loss": 0.6206, | |
| "step": 2028 | |
| }, | |
| { | |
| "epoch": 0.6023131672597865, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.876078751459856e-05, | |
| "loss": 0.5901, | |
| "step": 2031 | |
| }, | |
| { | |
| "epoch": 0.603202846975089, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 6.849460786452475e-05, | |
| "loss": 0.6007, | |
| "step": 2034 | |
| }, | |
| { | |
| "epoch": 0.6040925266903915, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 6.822867580350563e-05, | |
| "loss": 0.6419, | |
| "step": 2037 | |
| }, | |
| { | |
| "epoch": 0.604982206405694, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 6.79629934214015e-05, | |
| "loss": 0.6086, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.6058718861209964, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.769756280611046e-05, | |
| "loss": 0.6324, | |
| "step": 2043 | |
| }, | |
| { | |
| "epoch": 0.6067615658362989, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.743238604355219e-05, | |
| "loss": 0.5897, | |
| "step": 2046 | |
| }, | |
| { | |
| "epoch": 0.6076512455516014, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.716746521765131e-05, | |
| "loss": 0.598, | |
| "step": 2049 | |
| }, | |
| { | |
| "epoch": 0.608540925266904, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.690280241032116e-05, | |
| "loss": 0.6165, | |
| "step": 2052 | |
| }, | |
| { | |
| "epoch": 0.6094306049822064, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.663839970144751e-05, | |
| "loss": 0.607, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 0.6103202846975089, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 6.637425916887198e-05, | |
| "loss": 0.6024, | |
| "step": 2058 | |
| }, | |
| { | |
| "epoch": 0.6112099644128114, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 6.611038288837593e-05, | |
| "loss": 0.6074, | |
| "step": 2061 | |
| }, | |
| { | |
| "epoch": 0.6120996441281139, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 6.584677293366396e-05, | |
| "loss": 0.6156, | |
| "step": 2064 | |
| }, | |
| { | |
| "epoch": 0.6129893238434164, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 6.558343137634788e-05, | |
| "loss": 0.6304, | |
| "step": 2067 | |
| }, | |
| { | |
| "epoch": 0.6138790035587188, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 6.532036028593011e-05, | |
| "loss": 0.5998, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.6147686832740213, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 6.505756172978765e-05, | |
| "loss": 0.6092, | |
| "step": 2073 | |
| }, | |
| { | |
| "epoch": 0.6156583629893239, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.479503777315577e-05, | |
| "loss": 0.6055, | |
| "step": 2076 | |
| }, | |
| { | |
| "epoch": 0.6165480427046264, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.453279047911169e-05, | |
| "loss": 0.6009, | |
| "step": 2079 | |
| }, | |
| { | |
| "epoch": 0.6174377224199288, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.427082190855854e-05, | |
| "loss": 0.6202, | |
| "step": 2082 | |
| }, | |
| { | |
| "epoch": 0.6183274021352313, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 6.400913412020895e-05, | |
| "loss": 0.6211, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 0.6192170818505338, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.374772917056908e-05, | |
| "loss": 0.6151, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 0.6201067615658363, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 6.34866091139224e-05, | |
| "loss": 0.6111, | |
| "step": 2091 | |
| }, | |
| { | |
| "epoch": 0.6209964412811388, | |
| "grad_norm": 0.25, | |
| "learning_rate": 6.322577600231332e-05, | |
| "loss": 0.6109, | |
| "step": 2094 | |
| }, | |
| { | |
| "epoch": 0.6218861209964412, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.296523188553153e-05, | |
| "loss": 0.5829, | |
| "step": 2097 | |
| }, | |
| { | |
| "epoch": 0.6227758007117438, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.270497881109541e-05, | |
| "loss": 0.6002, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.6236654804270463, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.244501882423621e-05, | |
| "loss": 0.6095, | |
| "step": 2103 | |
| }, | |
| { | |
| "epoch": 0.6245551601423488, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 6.21853539678819e-05, | |
| "loss": 0.6049, | |
| "step": 2106 | |
| }, | |
| { | |
| "epoch": 0.6254448398576512, | |
| "grad_norm": 0.25, | |
| "learning_rate": 6.192598628264121e-05, | |
| "loss": 0.604, | |
| "step": 2109 | |
| }, | |
| { | |
| "epoch": 0.6263345195729537, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 6.166691780678743e-05, | |
| "loss": 0.6092, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 0.6272241992882562, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.140815057624248e-05, | |
| "loss": 0.6099, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 0.6281138790035588, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.114968662456093e-05, | |
| "loss": 0.5978, | |
| "step": 2118 | |
| }, | |
| { | |
| "epoch": 0.6290035587188612, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.089152798291398e-05, | |
| "loss": 0.5994, | |
| "step": 2121 | |
| }, | |
| { | |
| "epoch": 0.6298932384341637, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.063367668007356e-05, | |
| "loss": 0.5946, | |
| "step": 2124 | |
| }, | |
| { | |
| "epoch": 0.6307829181494662, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.0376134742396276e-05, | |
| "loss": 0.6131, | |
| "step": 2127 | |
| }, | |
| { | |
| "epoch": 0.6316725978647687, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.011890419380756e-05, | |
| "loss": 0.6259, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.6325622775800712, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.986198705578583e-05, | |
| "loss": 0.6079, | |
| "step": 2133 | |
| }, | |
| { | |
| "epoch": 0.6334519572953736, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 5.960538534734641e-05, | |
| "loss": 0.5792, | |
| "step": 2136 | |
| }, | |
| { | |
| "epoch": 0.6343416370106761, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 5.934910108502587e-05, | |
| "loss": 0.5913, | |
| "step": 2139 | |
| }, | |
| { | |
| "epoch": 0.6352313167259787, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 5.909313628286601e-05, | |
| "loss": 0.6215, | |
| "step": 2142 | |
| }, | |
| { | |
| "epoch": 0.6361209964412812, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.8837492952398234e-05, | |
| "loss": 0.5994, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 0.6370106761565836, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.8582173102627524e-05, | |
| "loss": 0.5984, | |
| "step": 2148 | |
| }, | |
| { | |
| "epoch": 0.6379003558718861, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 5.8327178740016744e-05, | |
| "loss": 0.612, | |
| "step": 2151 | |
| }, | |
| { | |
| "epoch": 0.6387900355871886, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.8072511868470945e-05, | |
| "loss": 0.596, | |
| "step": 2154 | |
| }, | |
| { | |
| "epoch": 0.6396797153024911, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.781817448932145e-05, | |
| "loss": 0.6327, | |
| "step": 2157 | |
| }, | |
| { | |
| "epoch": 0.6405693950177936, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.756416860131036e-05, | |
| "loss": 0.6162, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.641459074733096, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 5.731049620057457e-05, | |
| "loss": 0.6125, | |
| "step": 2163 | |
| }, | |
| { | |
| "epoch": 0.6423487544483986, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.705715928063031e-05, | |
| "loss": 0.6053, | |
| "step": 2166 | |
| }, | |
| { | |
| "epoch": 0.6432384341637011, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 5.6804159832357426e-05, | |
| "loss": 0.5802, | |
| "step": 2169 | |
| }, | |
| { | |
| "epoch": 0.6441281138790036, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 5.655149984398359e-05, | |
| "loss": 0.6088, | |
| "step": 2172 | |
| }, | |
| { | |
| "epoch": 0.645017793594306, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.629918130106886e-05, | |
| "loss": 0.6359, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.6459074733096085, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.6047206186489934e-05, | |
| "loss": 0.5961, | |
| "step": 2178 | |
| }, | |
| { | |
| "epoch": 0.646797153024911, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.5795576480424774e-05, | |
| "loss": 0.5898, | |
| "step": 2181 | |
| }, | |
| { | |
| "epoch": 0.6476868327402135, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 5.554429416033673e-05, | |
| "loss": 0.6315, | |
| "step": 2184 | |
| }, | |
| { | |
| "epoch": 0.6485765124555161, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 5.5293361200959314e-05, | |
| "loss": 0.6186, | |
| "step": 2187 | |
| }, | |
| { | |
| "epoch": 0.6494661921708185, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 5.504277957428052e-05, | |
| "loss": 0.6337, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.650355871886121, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.4792551249527314e-05, | |
| "loss": 0.6123, | |
| "step": 2193 | |
| }, | |
| { | |
| "epoch": 0.6512455516014235, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.454267819315015e-05, | |
| "loss": 0.6191, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 0.652135231316726, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.429316236880764e-05, | |
| "loss": 0.6093, | |
| "step": 2199 | |
| }, | |
| { | |
| "epoch": 0.6530249110320284, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 5.4044005737351044e-05, | |
| "loss": 0.6005, | |
| "step": 2202 | |
| }, | |
| { | |
| "epoch": 0.6539145907473309, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 5.379521025680878e-05, | |
| "loss": 0.6154, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 0.6548042704626335, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.3546777882371254e-05, | |
| "loss": 0.6227, | |
| "step": 2208 | |
| }, | |
| { | |
| "epoch": 0.655693950177936, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 5.329871056637524e-05, | |
| "loss": 0.6065, | |
| "step": 2211 | |
| }, | |
| { | |
| "epoch": 0.6565836298932385, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 5.305101025828863e-05, | |
| "loss": 0.6088, | |
| "step": 2214 | |
| }, | |
| { | |
| "epoch": 0.6574733096085409, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.280367890469529e-05, | |
| "loss": 0.5995, | |
| "step": 2217 | |
| }, | |
| { | |
| "epoch": 0.6583629893238434, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 5.255671844927944e-05, | |
| "loss": 0.6172, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.6592526690391459, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.231013083281067e-05, | |
| "loss": 0.603, | |
| "step": 2223 | |
| }, | |
| { | |
| "epoch": 0.6601423487544484, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.2063917993128554e-05, | |
| "loss": 0.6181, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 0.6610320284697508, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 5.1818081865127386e-05, | |
| "loss": 0.6015, | |
| "step": 2229 | |
| }, | |
| { | |
| "epoch": 0.6619217081850534, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 5.157262438074104e-05, | |
| "loss": 0.6155, | |
| "step": 2232 | |
| }, | |
| { | |
| "epoch": 0.6628113879003559, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 5.132754746892776e-05, | |
| "loss": 0.5928, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 0.6637010676156584, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 5.1082853055655076e-05, | |
| "loss": 0.6114, | |
| "step": 2238 | |
| }, | |
| { | |
| "epoch": 0.6645907473309609, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.0838543063884515e-05, | |
| "loss": 0.6169, | |
| "step": 2241 | |
| }, | |
| { | |
| "epoch": 0.6654804270462633, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.059461941355666e-05, | |
| "loss": 0.5962, | |
| "step": 2244 | |
| }, | |
| { | |
| "epoch": 0.6663701067615658, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 5.035108402157598e-05, | |
| "loss": 0.6391, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 0.6672597864768683, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 5.0107938801795695e-05, | |
| "loss": 0.6067, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6681494661921709, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.986518566500287e-05, | |
| "loss": 0.6251, | |
| "step": 2253 | |
| }, | |
| { | |
| "epoch": 0.6690391459074733, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 4.962282651890325e-05, | |
| "loss": 0.5799, | |
| "step": 2256 | |
| }, | |
| { | |
| "epoch": 0.6699288256227758, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 4.938086326810651e-05, | |
| "loss": 0.6, | |
| "step": 2259 | |
| }, | |
| { | |
| "epoch": 0.6708185053380783, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 4.913929781411098e-05, | |
| "loss": 0.5815, | |
| "step": 2262 | |
| }, | |
| { | |
| "epoch": 0.6717081850533808, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.889813205528895e-05, | |
| "loss": 0.5998, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 0.6725978647686833, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 4.865736788687164e-05, | |
| "loss": 0.6064, | |
| "step": 2268 | |
| }, | |
| { | |
| "epoch": 0.6734875444839857, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 4.8417007200934294e-05, | |
| "loss": 0.5976, | |
| "step": 2271 | |
| }, | |
| { | |
| "epoch": 0.6743772241992882, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.8177051886381344e-05, | |
| "loss": 0.5906, | |
| "step": 2274 | |
| }, | |
| { | |
| "epoch": 0.6752669039145908, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 4.793750382893151e-05, | |
| "loss": 0.6096, | |
| "step": 2277 | |
| }, | |
| { | |
| "epoch": 0.6761565836298933, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.769836491110314e-05, | |
| "loss": 0.6067, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.6770462633451957, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 4.74596370121993e-05, | |
| "loss": 0.6047, | |
| "step": 2283 | |
| }, | |
| { | |
| "epoch": 0.6779359430604982, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 4.7221322008292915e-05, | |
| "loss": 0.5876, | |
| "step": 2286 | |
| }, | |
| { | |
| "epoch": 0.6788256227758007, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 4.698342177221219e-05, | |
| "loss": 0.6129, | |
| "step": 2289 | |
| }, | |
| { | |
| "epoch": 0.6797153024911032, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.674593817352575e-05, | |
| "loss": 0.6128, | |
| "step": 2292 | |
| }, | |
| { | |
| "epoch": 0.6806049822064056, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 4.650887307852818e-05, | |
| "loss": 0.6152, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 0.6814946619217082, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.627222835022502e-05, | |
| "loss": 0.6167, | |
| "step": 2298 | |
| }, | |
| { | |
| "epoch": 0.6823843416370107, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 4.603600584831844e-05, | |
| "loss": 0.6155, | |
| "step": 2301 | |
| }, | |
| { | |
| "epoch": 0.6832740213523132, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 4.580020742919246e-05, | |
| "loss": 0.6212, | |
| "step": 2304 | |
| }, | |
| { | |
| "epoch": 0.6841637010676157, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 4.556483494589836e-05, | |
| "loss": 0.6115, | |
| "step": 2307 | |
| }, | |
| { | |
| "epoch": 0.6850533807829181, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 4.532989024814015e-05, | |
| "loss": 0.5939, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.6859430604982206, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.5095375182260016e-05, | |
| "loss": 0.5834, | |
| "step": 2313 | |
| }, | |
| { | |
| "epoch": 0.6868327402135231, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.486129159122393e-05, | |
| "loss": 0.5859, | |
| "step": 2316 | |
| }, | |
| { | |
| "epoch": 0.6877224199288257, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 4.462764131460694e-05, | |
| "loss": 0.6076, | |
| "step": 2319 | |
| }, | |
| { | |
| "epoch": 0.6886120996441281, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.439442618857891e-05, | |
| "loss": 0.5954, | |
| "step": 2322 | |
| }, | |
| { | |
| "epoch": 0.6895017793594306, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.416164804589005e-05, | |
| "loss": 0.6072, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.6903914590747331, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 4.39293087158564e-05, | |
| "loss": 0.6083, | |
| "step": 2328 | |
| }, | |
| { | |
| "epoch": 0.6912811387900356, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 4.369741002434556e-05, | |
| "loss": 0.5948, | |
| "step": 2331 | |
| }, | |
| { | |
| "epoch": 0.6921708185053381, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.346595379376232e-05, | |
| "loss": 0.6148, | |
| "step": 2334 | |
| }, | |
| { | |
| "epoch": 0.6930604982206405, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 4.323494184303435e-05, | |
| "loss": 0.6134, | |
| "step": 2337 | |
| }, | |
| { | |
| "epoch": 0.693950177935943, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 4.3004375987597946e-05, | |
| "loss": 0.5801, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.6948398576512456, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.277425803938356e-05, | |
| "loss": 0.615, | |
| "step": 2343 | |
| }, | |
| { | |
| "epoch": 0.6957295373665481, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 4.254458980680188e-05, | |
| "loss": 0.6239, | |
| "step": 2346 | |
| }, | |
| { | |
| "epoch": 0.6966192170818505, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 4.2315373094729316e-05, | |
| "loss": 0.5997, | |
| "step": 2349 | |
| }, | |
| { | |
| "epoch": 0.697508896797153, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 4.2086609704494015e-05, | |
| "loss": 0.5897, | |
| "step": 2352 | |
| }, | |
| { | |
| "epoch": 0.6983985765124555, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.1858301433861566e-05, | |
| "loss": 0.5926, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 0.699288256227758, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.163045007702104e-05, | |
| "loss": 0.5991, | |
| "step": 2358 | |
| }, | |
| { | |
| "epoch": 0.7001779359430605, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.14030574245708e-05, | |
| "loss": 0.6193, | |
| "step": 2361 | |
| }, | |
| { | |
| "epoch": 0.701067615658363, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 4.117612526350428e-05, | |
| "loss": 0.6146, | |
| "step": 2364 | |
| }, | |
| { | |
| "epoch": 0.7019572953736655, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.09496553771963e-05, | |
| "loss": 0.6032, | |
| "step": 2367 | |
| }, | |
| { | |
| "epoch": 0.702846975088968, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.0723649545388575e-05, | |
| "loss": 0.5999, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.7037366548042705, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 4.0498109544176245e-05, | |
| "loss": 0.5979, | |
| "step": 2373 | |
| }, | |
| { | |
| "epoch": 0.7046263345195729, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.0273037145993454e-05, | |
| "loss": 0.6016, | |
| "step": 2376 | |
| }, | |
| { | |
| "epoch": 0.7055160142348754, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.0048434119599765e-05, | |
| "loss": 0.6017, | |
| "step": 2379 | |
| }, | |
| { | |
| "epoch": 0.7064056939501779, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.982430223006613e-05, | |
| "loss": 0.5984, | |
| "step": 2382 | |
| }, | |
| { | |
| "epoch": 0.7072953736654805, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 3.960064323876093e-05, | |
| "loss": 0.5982, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 0.708185053380783, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.937745890333623e-05, | |
| "loss": 0.5984, | |
| "step": 2388 | |
| }, | |
| { | |
| "epoch": 0.7090747330960854, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.915475097771396e-05, | |
| "loss": 0.6207, | |
| "step": 2391 | |
| }, | |
| { | |
| "epoch": 0.7099644128113879, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 3.8932521212072206e-05, | |
| "loss": 0.6029, | |
| "step": 2394 | |
| }, | |
| { | |
| "epoch": 0.7108540925266904, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.871077135283123e-05, | |
| "loss": 0.5887, | |
| "step": 2397 | |
| }, | |
| { | |
| "epoch": 0.7117437722419929, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.8489503142640016e-05, | |
| "loss": 0.6058, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.7126334519572953, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 3.826871832036242e-05, | |
| "loss": 0.606, | |
| "step": 2403 | |
| }, | |
| { | |
| "epoch": 0.7135231316725978, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.804841862106347e-05, | |
| "loss": 0.5846, | |
| "step": 2406 | |
| }, | |
| { | |
| "epoch": 0.7144128113879004, | |
| "grad_norm": 0.287109375, | |
| "learning_rate": 3.782860577599585e-05, | |
| "loss": 0.6075, | |
| "step": 2409 | |
| }, | |
| { | |
| "epoch": 0.7153024911032029, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 3.7609281512586203e-05, | |
| "loss": 0.6038, | |
| "step": 2412 | |
| }, | |
| { | |
| "epoch": 0.7161921708185054, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.739044755442162e-05, | |
| "loss": 0.6201, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 0.7170818505338078, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.717210562123613e-05, | |
| "loss": 0.6062, | |
| "step": 2418 | |
| }, | |
| { | |
| "epoch": 0.7179715302491103, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 3.695425742889698e-05, | |
| "loss": 0.6108, | |
| "step": 2421 | |
| }, | |
| { | |
| "epoch": 0.7188612099644128, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 3.6736904689391417e-05, | |
| "loss": 0.6307, | |
| "step": 2424 | |
| }, | |
| { | |
| "epoch": 0.7197508896797153, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 3.6520049110813035e-05, | |
| "loss": 0.6057, | |
| "step": 2427 | |
| }, | |
| { | |
| "epoch": 0.7206405693950177, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 3.6303692397348455e-05, | |
| "loss": 0.6262, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.7215302491103203, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.6087836249263875e-05, | |
| "loss": 0.599, | |
| "step": 2433 | |
| }, | |
| { | |
| "epoch": 0.7224199288256228, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.58724823628918e-05, | |
| "loss": 0.5879, | |
| "step": 2436 | |
| }, | |
| { | |
| "epoch": 0.7233096085409253, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 3.5657632430617635e-05, | |
| "loss": 0.6095, | |
| "step": 2439 | |
| }, | |
| { | |
| "epoch": 0.7241992882562278, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.5443288140866316e-05, | |
| "loss": 0.593, | |
| "step": 2442 | |
| }, | |
| { | |
| "epoch": 0.7250889679715302, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.522945117808929e-05, | |
| "loss": 0.5932, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 0.7259786476868327, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 3.501612322275086e-05, | |
| "loss": 0.6149, | |
| "step": 2448 | |
| }, | |
| { | |
| "epoch": 0.7268683274021353, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 3.48033059513155e-05, | |
| "loss": 0.591, | |
| "step": 2451 | |
| }, | |
| { | |
| "epoch": 0.7277580071174378, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.45910010362342e-05, | |
| "loss": 0.5986, | |
| "step": 2454 | |
| }, | |
| { | |
| "epoch": 0.7286476868327402, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 3.437921014593167e-05, | |
| "loss": 0.5983, | |
| "step": 2457 | |
| }, | |
| { | |
| "epoch": 0.7295373665480427, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 3.416793494479308e-05, | |
| "loss": 0.6305, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.7304270462633452, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.3957177093150915e-05, | |
| "loss": 0.6212, | |
| "step": 2463 | |
| }, | |
| { | |
| "epoch": 0.7313167259786477, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 3.374693824727204e-05, | |
| "loss": 0.6132, | |
| "step": 2466 | |
| }, | |
| { | |
| "epoch": 0.7322064056939501, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.353722005934463e-05, | |
| "loss": 0.5886, | |
| "step": 2469 | |
| }, | |
| { | |
| "epoch": 0.7330960854092526, | |
| "grad_norm": 0.296875, | |
| "learning_rate": 3.332802417746527e-05, | |
| "loss": 0.6087, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 0.7339857651245552, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.311935224562591e-05, | |
| "loss": 0.6045, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.7348754448398577, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.291120590370091e-05, | |
| "loss": 0.5976, | |
| "step": 2478 | |
| }, | |
| { | |
| "epoch": 0.7357651245551602, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.270358678743434e-05, | |
| "loss": 0.6191, | |
| "step": 2481 | |
| }, | |
| { | |
| "epoch": 0.7366548042704626, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 3.249649652842687e-05, | |
| "loss": 0.5864, | |
| "step": 2484 | |
| }, | |
| { | |
| "epoch": 0.7375444839857651, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 3.228993675412315e-05, | |
| "loss": 0.6063, | |
| "step": 2487 | |
| }, | |
| { | |
| "epoch": 0.7384341637010676, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.20839090877989e-05, | |
| "loss": 0.5998, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.7393238434163701, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 3.187841514854829e-05, | |
| "loss": 0.5807, | |
| "step": 2493 | |
| }, | |
| { | |
| "epoch": 0.7402135231316725, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 3.1673456551271086e-05, | |
| "loss": 0.5983, | |
| "step": 2496 | |
| }, | |
| { | |
| "epoch": 0.7411032028469751, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.1469034906659946e-05, | |
| "loss": 0.6053, | |
| "step": 2499 | |
| }, | |
| { | |
| "epoch": 0.7419928825622776, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 3.126515182118793e-05, | |
| "loss": 0.5994, | |
| "step": 2502 | |
| }, | |
| { | |
| "epoch": 0.7428825622775801, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 3.106180889709567e-05, | |
| "loss": 0.5969, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 0.7437722419928826, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.0859007732378896e-05, | |
| "loss": 0.5936, | |
| "step": 2508 | |
| }, | |
| { | |
| "epoch": 0.744661921708185, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.065674992077584e-05, | |
| "loss": 0.5717, | |
| "step": 2511 | |
| }, | |
| { | |
| "epoch": 0.7455516014234875, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.0455037051754777e-05, | |
| "loss": 0.6061, | |
| "step": 2514 | |
| }, | |
| { | |
| "epoch": 0.74644128113879, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.0253870710501475e-05, | |
| "loss": 0.5914, | |
| "step": 2517 | |
| }, | |
| { | |
| "epoch": 0.7473309608540926, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.005325247790668e-05, | |
| "loss": 0.6067, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.748220640569395, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 2.9853183930553853e-05, | |
| "loss": 0.5909, | |
| "step": 2523 | |
| }, | |
| { | |
| "epoch": 0.7491103202846975, | |
| "grad_norm": 0.25, | |
| "learning_rate": 2.965366664070661e-05, | |
| "loss": 0.5847, | |
| "step": 2526 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.9454702176296423e-05, | |
| "loss": 0.5907, | |
| "step": 2529 | |
| }, | |
| { | |
| "epoch": 0.7508896797153025, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.925629210091043e-05, | |
| "loss": 0.606, | |
| "step": 2532 | |
| }, | |
| { | |
| "epoch": 0.751779359430605, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 2.9058437973778896e-05, | |
| "loss": 0.6055, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 0.7526690391459074, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.886114134976322e-05, | |
| "loss": 0.5993, | |
| "step": 2538 | |
| }, | |
| { | |
| "epoch": 0.75355871886121, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 2.866440377934352e-05, | |
| "loss": 0.6098, | |
| "step": 2541 | |
| }, | |
| { | |
| "epoch": 0.7544483985765125, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.8468226808606522e-05, | |
| "loss": 0.584, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 0.755338078291815, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.827261197923341e-05, | |
| "loss": 0.5949, | |
| "step": 2547 | |
| }, | |
| { | |
| "epoch": 0.7562277580071174, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 2.8077560828487748e-05, | |
| "loss": 0.5698, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7571174377224199, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.7883074889203363e-05, | |
| "loss": 0.612, | |
| "step": 2553 | |
| }, | |
| { | |
| "epoch": 0.7580071174377224, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.7689155689772217e-05, | |
| "loss": 0.5951, | |
| "step": 2556 | |
| }, | |
| { | |
| "epoch": 0.7588967971530249, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.7495804754132602e-05, | |
| "loss": 0.5841, | |
| "step": 2559 | |
| }, | |
| { | |
| "epoch": 0.7597864768683275, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.7303023601756928e-05, | |
| "loss": 0.5978, | |
| "step": 2562 | |
| }, | |
| { | |
| "epoch": 0.7606761565836299, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 2.711081374763993e-05, | |
| "loss": 0.5994, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 0.7615658362989324, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.6919176702286698e-05, | |
| "loss": 0.6014, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 0.7624555160142349, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 2.6728113971700908e-05, | |
| "loss": 0.5958, | |
| "step": 2571 | |
| }, | |
| { | |
| "epoch": 0.7633451957295374, | |
| "grad_norm": 0.28125, | |
| "learning_rate": 2.653762705737287e-05, | |
| "loss": 0.6242, | |
| "step": 2574 | |
| }, | |
| { | |
| "epoch": 0.7642348754448398, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.634771745626772e-05, | |
| "loss": 0.616, | |
| "step": 2577 | |
| }, | |
| { | |
| "epoch": 0.7651245551601423, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 2.6158386660813806e-05, | |
| "loss": 0.5959, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.7660142348754448, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.5969636158890775e-05, | |
| "loss": 0.5971, | |
| "step": 2583 | |
| }, | |
| { | |
| "epoch": 0.7669039145907474, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.5781467433817973e-05, | |
| "loss": 0.593, | |
| "step": 2586 | |
| }, | |
| { | |
| "epoch": 0.7677935943060499, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.5593881964342857e-05, | |
| "loss": 0.5841, | |
| "step": 2589 | |
| }, | |
| { | |
| "epoch": 0.7686832740213523, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 2.5406881224629174e-05, | |
| "loss": 0.6111, | |
| "step": 2592 | |
| }, | |
| { | |
| "epoch": 0.7695729537366548, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.5220466684245646e-05, | |
| "loss": 0.5758, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 0.7704626334519573, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.5034639808154114e-05, | |
| "loss": 0.6276, | |
| "step": 2598 | |
| }, | |
| { | |
| "epoch": 0.7713523131672598, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.4849402056698334e-05, | |
| "loss": 0.6062, | |
| "step": 2601 | |
| }, | |
| { | |
| "epoch": 0.7722419928825622, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.4664754885592268e-05, | |
| "loss": 0.5881, | |
| "step": 2604 | |
| }, | |
| { | |
| "epoch": 0.7731316725978647, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.4480699745908707e-05, | |
| "loss": 0.6124, | |
| "step": 2607 | |
| }, | |
| { | |
| "epoch": 0.7740213523131673, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.4297238084067985e-05, | |
| "loss": 0.5779, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.7749110320284698, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.4114371341826415e-05, | |
| "loss": 0.6019, | |
| "step": 2613 | |
| }, | |
| { | |
| "epoch": 0.7758007117437722, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.3932100956265148e-05, | |
| "loss": 0.6087, | |
| "step": 2616 | |
| }, | |
| { | |
| "epoch": 0.7766903914590747, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.375042835977872e-05, | |
| "loss": 0.5983, | |
| "step": 2619 | |
| }, | |
| { | |
| "epoch": 0.7775800711743772, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 2.3569354980063906e-05, | |
| "loss": 0.6024, | |
| "step": 2622 | |
| }, | |
| { | |
| "epoch": 0.7784697508896797, | |
| "grad_norm": 0.25, | |
| "learning_rate": 2.3388882240108423e-05, | |
| "loss": 0.6039, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.7793594306049823, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 2.3209011558179826e-05, | |
| "loss": 0.5958, | |
| "step": 2628 | |
| }, | |
| { | |
| "epoch": 0.7802491103202847, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.3029744347814365e-05, | |
| "loss": 0.5979, | |
| "step": 2631 | |
| }, | |
| { | |
| "epoch": 0.7811387900355872, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 2.2851082017805703e-05, | |
| "loss": 0.5918, | |
| "step": 2634 | |
| }, | |
| { | |
| "epoch": 0.7820284697508897, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.2673025972194106e-05, | |
| "loss": 0.5906, | |
| "step": 2637 | |
| }, | |
| { | |
| "epoch": 0.7829181494661922, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.2495577610255203e-05, | |
| "loss": 0.5857, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.7838078291814946, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.2318738326489074e-05, | |
| "loss": 0.602, | |
| "step": 2643 | |
| }, | |
| { | |
| "epoch": 0.7846975088967971, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.2142509510609277e-05, | |
| "loss": 0.5846, | |
| "step": 2646 | |
| }, | |
| { | |
| "epoch": 0.7855871886120996, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.196689254753196e-05, | |
| "loss": 0.5983, | |
| "step": 2649 | |
| }, | |
| { | |
| "epoch": 0.7864768683274022, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.179188881736498e-05, | |
| "loss": 0.5753, | |
| "step": 2652 | |
| }, | |
| { | |
| "epoch": 0.7873665480427047, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.1617499695396924e-05, | |
| "loss": 0.605, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 0.7882562277580071, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.1443726552086528e-05, | |
| "loss": 0.5982, | |
| "step": 2658 | |
| }, | |
| { | |
| "epoch": 0.7891459074733096, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 2.1270570753051668e-05, | |
| "loss": 0.5972, | |
| "step": 2661 | |
| }, | |
| { | |
| "epoch": 0.7900355871886121, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 2.109803365905879e-05, | |
| "loss": 0.5869, | |
| "step": 2664 | |
| }, | |
| { | |
| "epoch": 0.7909252669039146, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 2.0926116626012205e-05, | |
| "loss": 0.5984, | |
| "step": 2667 | |
| }, | |
| { | |
| "epoch": 0.791814946619217, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.0754821004943336e-05, | |
| "loss": 0.6054, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.7927046263345195, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.0584148142000225e-05, | |
| "loss": 0.5809, | |
| "step": 2673 | |
| }, | |
| { | |
| "epoch": 0.7935943060498221, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 2.0414099378436813e-05, | |
| "loss": 0.6019, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 0.7944839857651246, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.0244676050602572e-05, | |
| "loss": 0.5837, | |
| "step": 2679 | |
| }, | |
| { | |
| "epoch": 0.7953736654804271, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 2.0075879489931847e-05, | |
| "loss": 0.6226, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 0.7962633451957295, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.990771102293344e-05, | |
| "loss": 0.6255, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 0.797153024911032, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.9740171971180278e-05, | |
| "loss": 0.5958, | |
| "step": 2688 | |
| }, | |
| { | |
| "epoch": 0.7980427046263345, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.9573263651298836e-05, | |
| "loss": 0.5902, | |
| "step": 2691 | |
| }, | |
| { | |
| "epoch": 0.798932384341637, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.940698737495904e-05, | |
| "loss": 0.6078, | |
| "step": 2694 | |
| }, | |
| { | |
| "epoch": 0.7998220640569395, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.9241344448863696e-05, | |
| "loss": 0.6009, | |
| "step": 2697 | |
| }, | |
| { | |
| "epoch": 0.800711743772242, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 1.9076336174738473e-05, | |
| "loss": 0.5657, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.8016014234875445, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.891196384932139e-05, | |
| "loss": 0.5824, | |
| "step": 2703 | |
| }, | |
| { | |
| "epoch": 0.802491103202847, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.8748228764352914e-05, | |
| "loss": 0.5945, | |
| "step": 2706 | |
| }, | |
| { | |
| "epoch": 0.8033807829181495, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.858513220656567e-05, | |
| "loss": 0.5806, | |
| "step": 2709 | |
| }, | |
| { | |
| "epoch": 0.8042704626334519, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.8422675457674254e-05, | |
| "loss": 0.5963, | |
| "step": 2712 | |
| }, | |
| { | |
| "epoch": 0.8051601423487544, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.8260859794365338e-05, | |
| "loss": 0.5849, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 0.806049822064057, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 1.809968648828748e-05, | |
| "loss": 0.6066, | |
| "step": 2718 | |
| }, | |
| { | |
| "epoch": 0.8069395017793595, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.7939156806041203e-05, | |
| "loss": 0.5969, | |
| "step": 2721 | |
| }, | |
| { | |
| "epoch": 0.8078291814946619, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.777927200916907e-05, | |
| "loss": 0.5832, | |
| "step": 2724 | |
| }, | |
| { | |
| "epoch": 0.8087188612099644, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.762003335414566e-05, | |
| "loss": 0.5965, | |
| "step": 2727 | |
| }, | |
| { | |
| "epoch": 0.8096085409252669, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.7461442092367862e-05, | |
| "loss": 0.5792, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.8104982206405694, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 1.7303499470144846e-05, | |
| "loss": 0.5943, | |
| "step": 2733 | |
| }, | |
| { | |
| "epoch": 0.8113879003558719, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.7146206728688463e-05, | |
| "loss": 0.6036, | |
| "step": 2736 | |
| }, | |
| { | |
| "epoch": 0.8122775800711743, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.6989565104103312e-05, | |
| "loss": 0.5755, | |
| "step": 2739 | |
| }, | |
| { | |
| "epoch": 0.8131672597864769, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.6833575827377134e-05, | |
| "loss": 0.6149, | |
| "step": 2742 | |
| }, | |
| { | |
| "epoch": 0.8140569395017794, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.6678240124371157e-05, | |
| "loss": 0.5944, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 0.8149466192170819, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.6523559215810337e-05, | |
| "loss": 0.5925, | |
| "step": 2748 | |
| }, | |
| { | |
| "epoch": 0.8158362989323843, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.636953431727395e-05, | |
| "loss": 0.5936, | |
| "step": 2751 | |
| }, | |
| { | |
| "epoch": 0.8167259786476868, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.6216166639185803e-05, | |
| "loss": 0.5973, | |
| "step": 2754 | |
| }, | |
| { | |
| "epoch": 0.8176156583629893, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.6063457386805004e-05, | |
| "loss": 0.6125, | |
| "step": 2757 | |
| }, | |
| { | |
| "epoch": 0.8185053380782918, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.5911407760216235e-05, | |
| "loss": 0.6036, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.8193950177935944, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.576001895432042e-05, | |
| "loss": 0.597, | |
| "step": 2763 | |
| }, | |
| { | |
| "epoch": 0.8202846975088968, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.5609292158825438e-05, | |
| "loss": 0.5872, | |
| "step": 2766 | |
| }, | |
| { | |
| "epoch": 0.8211743772241993, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.545922855823656e-05, | |
| "loss": 0.6095, | |
| "step": 2769 | |
| }, | |
| { | |
| "epoch": 0.8220640569395018, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.530982933184737e-05, | |
| "loss": 0.603, | |
| "step": 2772 | |
| }, | |
| { | |
| "epoch": 0.8229537366548043, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.5161095653730273e-05, | |
| "loss": 0.5938, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.8238434163701067, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 1.5013028692727481e-05, | |
| "loss": 0.6032, | |
| "step": 2778 | |
| }, | |
| { | |
| "epoch": 0.8247330960854092, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.4865629612441656e-05, | |
| "loss": 0.5877, | |
| "step": 2781 | |
| }, | |
| { | |
| "epoch": 0.8256227758007118, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.471889957122684e-05, | |
| "loss": 0.6057, | |
| "step": 2784 | |
| }, | |
| { | |
| "epoch": 0.8265124555160143, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.457283972217941e-05, | |
| "loss": 0.6011, | |
| "step": 2787 | |
| }, | |
| { | |
| "epoch": 0.8274021352313167, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.4427451213128873e-05, | |
| "loss": 0.6009, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.8282918149466192, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 1.4282735186629014e-05, | |
| "loss": 0.6263, | |
| "step": 2793 | |
| }, | |
| { | |
| "epoch": 0.8291814946619217, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.4138692779948748e-05, | |
| "loss": 0.5978, | |
| "step": 2796 | |
| }, | |
| { | |
| "epoch": 0.8300711743772242, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.3995325125063274e-05, | |
| "loss": 0.5968, | |
| "step": 2799 | |
| }, | |
| { | |
| "epoch": 0.8309608540925267, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.3852633348645262e-05, | |
| "loss": 0.6074, | |
| "step": 2802 | |
| }, | |
| { | |
| "epoch": 0.8318505338078291, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.3710618572055767e-05, | |
| "loss": 0.608, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 0.8327402135231317, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.3569281911335684e-05, | |
| "loss": 0.5896, | |
| "step": 2808 | |
| }, | |
| { | |
| "epoch": 0.8336298932384342, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.3428624477196761e-05, | |
| "loss": 0.6042, | |
| "step": 2811 | |
| }, | |
| { | |
| "epoch": 0.8345195729537367, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.328864737501302e-05, | |
| "loss": 0.6092, | |
| "step": 2814 | |
| }, | |
| { | |
| "epoch": 0.8354092526690391, | |
| "grad_norm": 0.25, | |
| "learning_rate": 1.3149351704811962e-05, | |
| "loss": 0.6081, | |
| "step": 2817 | |
| }, | |
| { | |
| "epoch": 0.8362989323843416, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 1.3010738561265979e-05, | |
| "loss": 0.5918, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.8371886120996441, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.2872809033683798e-05, | |
| "loss": 0.595, | |
| "step": 2823 | |
| }, | |
| { | |
| "epoch": 0.8380782918149466, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 1.2735564206001749e-05, | |
| "loss": 0.5856, | |
| "step": 2826 | |
| }, | |
| { | |
| "epoch": 0.8389679715302492, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 1.2599005156775512e-05, | |
| "loss": 0.5715, | |
| "step": 2829 | |
| }, | |
| { | |
| "epoch": 0.8398576512455516, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.2463132959171341e-05, | |
| "loss": 0.607, | |
| "step": 2832 | |
| }, | |
| { | |
| "epoch": 0.8407473309608541, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.2327948680957924e-05, | |
| "loss": 0.5798, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 0.8416370106761566, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.2193453384497722e-05, | |
| "loss": 0.6194, | |
| "step": 2838 | |
| }, | |
| { | |
| "epoch": 0.8425266903914591, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.205964812673881e-05, | |
| "loss": 0.597, | |
| "step": 2841 | |
| }, | |
| { | |
| "epoch": 0.8434163701067615, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 1.192653395920652e-05, | |
| "loss": 0.6119, | |
| "step": 2844 | |
| }, | |
| { | |
| "epoch": 0.844306049822064, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.179411192799511e-05, | |
| "loss": 0.5891, | |
| "step": 2847 | |
| }, | |
| { | |
| "epoch": 0.8451957295373665, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.1662383073759685e-05, | |
| "loss": 0.6034, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.8460854092526691, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.1531348431707823e-05, | |
| "loss": 0.5686, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 0.8469750889679716, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.1401009031591658e-05, | |
| "loss": 0.5882, | |
| "step": 2856 | |
| }, | |
| { | |
| "epoch": 0.847864768683274, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.1271365897699615e-05, | |
| "loss": 0.5835, | |
| "step": 2859 | |
| }, | |
| { | |
| "epoch": 0.8487544483985765, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.114242004884839e-05, | |
| "loss": 0.6015, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 0.849644128113879, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.1014172498375086e-05, | |
| "loss": 0.6059, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 0.8505338078291815, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.088662425412903e-05, | |
| "loss": 0.5979, | |
| "step": 2868 | |
| }, | |
| { | |
| "epoch": 0.8514234875444839, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.0759776318464043e-05, | |
| "loss": 0.6005, | |
| "step": 2871 | |
| }, | |
| { | |
| "epoch": 0.8523131672597865, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.0633629688230452e-05, | |
| "loss": 0.595, | |
| "step": 2874 | |
| }, | |
| { | |
| "epoch": 0.853202846975089, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.0508185354767264e-05, | |
| "loss": 0.5866, | |
| "step": 2877 | |
| }, | |
| { | |
| "epoch": 0.8540925266903915, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.0383444303894452e-05, | |
| "loss": 0.6049, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.854982206405694, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.0259407515905094e-05, | |
| "loss": 0.6029, | |
| "step": 2883 | |
| }, | |
| { | |
| "epoch": 0.8558718861209964, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.0136075965557811e-05, | |
| "loss": 0.6028, | |
| "step": 2886 | |
| }, | |
| { | |
| "epoch": 0.8567615658362989, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 1.0013450622068921e-05, | |
| "loss": 0.6034, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 0.8576512455516014, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 9.891532449105045e-06, | |
| "loss": 0.6008, | |
| "step": 2892 | |
| }, | |
| { | |
| "epoch": 0.858540925266904, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 9.770322404775323e-06, | |
| "loss": 0.5886, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 0.8594306049822064, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 9.649821441623986e-06, | |
| "loss": 0.5991, | |
| "step": 2898 | |
| }, | |
| { | |
| "epoch": 0.8603202846975089, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 9.530030506622934e-06, | |
| "loss": 0.6189, | |
| "step": 2901 | |
| }, | |
| { | |
| "epoch": 0.8612099644128114, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 9.410950541164143e-06, | |
| "loss": 0.6034, | |
| "step": 2904 | |
| }, | |
| { | |
| "epoch": 0.8620996441281139, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 9.292582481052403e-06, | |
| "loss": 0.5953, | |
| "step": 2907 | |
| }, | |
| { | |
| "epoch": 0.8629893238434164, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.174927256497844e-06, | |
| "loss": 0.5974, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.8638790035587188, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 9.05798579210878e-06, | |
| "loss": 0.6074, | |
| "step": 2913 | |
| }, | |
| { | |
| "epoch": 0.8647686832740213, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 8.941759006884265e-06, | |
| "loss": 0.5955, | |
| "step": 2916 | |
| }, | |
| { | |
| "epoch": 0.8656583629893239, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.826247814206967e-06, | |
| "loss": 0.6182, | |
| "step": 2919 | |
| }, | |
| { | |
| "epoch": 0.8665480427046264, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 8.711453121836066e-06, | |
| "loss": 0.5978, | |
| "step": 2922 | |
| }, | |
| { | |
| "epoch": 0.8674377224199288, | |
| "grad_norm": 0.25, | |
| "learning_rate": 8.597375831899913e-06, | |
| "loss": 0.5886, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.8683274021352313, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 8.484016840889176e-06, | |
| "loss": 0.6046, | |
| "step": 2928 | |
| }, | |
| { | |
| "epoch": 0.8692170818505338, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 8.371377039649586e-06, | |
| "loss": 0.6104, | |
| "step": 2931 | |
| }, | |
| { | |
| "epoch": 0.8701067615658363, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 8.259457313375096e-06, | |
| "loss": 0.6054, | |
| "step": 2934 | |
| }, | |
| { | |
| "epoch": 0.8709964412811388, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 8.14825854160085e-06, | |
| "loss": 0.5899, | |
| "step": 2937 | |
| }, | |
| { | |
| "epoch": 0.8718861209964412, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 8.037781598196225e-06, | |
| "loss": 0.5991, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.8727758007117438, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.928027351358114e-06, | |
| "loss": 0.5856, | |
| "step": 2943 | |
| }, | |
| { | |
| "epoch": 0.8736654804270463, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 7.818996663603917e-06, | |
| "loss": 0.6008, | |
| "step": 2946 | |
| }, | |
| { | |
| "epoch": 0.8745551601423488, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 7.71069039176493e-06, | |
| "loss": 0.5707, | |
| "step": 2949 | |
| }, | |
| { | |
| "epoch": 0.8754448398576512, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 7.603109386979501e-06, | |
| "loss": 0.588, | |
| "step": 2952 | |
| }, | |
| { | |
| "epoch": 0.8763345195729537, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 7.496254494686339e-06, | |
| "loss": 0.6, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 0.8772241992882562, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 7.390126554617982e-06, | |
| "loss": 0.5762, | |
| "step": 2958 | |
| }, | |
| { | |
| "epoch": 0.8781138790035588, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.284726400794073e-06, | |
| "loss": 0.5991, | |
| "step": 2961 | |
| }, | |
| { | |
| "epoch": 0.8790035587188612, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 7.180054861514885e-06, | |
| "loss": 0.5988, | |
| "step": 2964 | |
| }, | |
| { | |
| "epoch": 0.8798932384341637, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 7.076112759354736e-06, | |
| "loss": 0.5755, | |
| "step": 2967 | |
| }, | |
| { | |
| "epoch": 0.8807829181494662, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.972900911155655e-06, | |
| "loss": 0.609, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.8816725978647687, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.8704201280207935e-06, | |
| "loss": 0.5757, | |
| "step": 2973 | |
| }, | |
| { | |
| "epoch": 0.8825622775800712, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.7686712153081645e-06, | |
| "loss": 0.6276, | |
| "step": 2976 | |
| }, | |
| { | |
| "epoch": 0.8834519572953736, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.667654972624315e-06, | |
| "loss": 0.6059, | |
| "step": 2979 | |
| }, | |
| { | |
| "epoch": 0.8843416370106761, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 6.567372193817966e-06, | |
| "loss": 0.5917, | |
| "step": 2982 | |
| }, | |
| { | |
| "epoch": 0.8852313167259787, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.467823666973871e-06, | |
| "loss": 0.5702, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 0.8861209964412812, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.369010174406531e-06, | |
| "loss": 0.607, | |
| "step": 2988 | |
| }, | |
| { | |
| "epoch": 0.8870106761565836, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 6.270932492654125e-06, | |
| "loss": 0.5965, | |
| "step": 2991 | |
| }, | |
| { | |
| "epoch": 0.8879003558718861, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.173591392472333e-06, | |
| "loss": 0.587, | |
| "step": 2994 | |
| }, | |
| { | |
| "epoch": 0.8887900355871886, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 6.076987638828335e-06, | |
| "loss": 0.6043, | |
| "step": 2997 | |
| }, | |
| { | |
| "epoch": 0.8896797153024911, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 5.981121990894789e-06, | |
| "loss": 0.5967, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8905693950177936, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 5.885995202043848e-06, | |
| "loss": 0.584, | |
| "step": 3003 | |
| }, | |
| { | |
| "epoch": 0.891459074733096, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 5.791608019841244e-06, | |
| "loss": 0.5836, | |
| "step": 3006 | |
| }, | |
| { | |
| "epoch": 0.8923487544483986, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 5.697961186040435e-06, | |
| "loss": 0.6029, | |
| "step": 3009 | |
| }, | |
| { | |
| "epoch": 0.8932384341637011, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 5.605055436576745e-06, | |
| "loss": 0.588, | |
| "step": 3012 | |
| }, | |
| { | |
| "epoch": 0.8941281138790036, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 5.51289150156159e-06, | |
| "loss": 0.596, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 0.895017793594306, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.421470105276749e-06, | |
| "loss": 0.5913, | |
| "step": 3018 | |
| }, | |
| { | |
| "epoch": 0.8959074733096085, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 5.33079196616868e-06, | |
| "loss": 0.5799, | |
| "step": 3021 | |
| }, | |
| { | |
| "epoch": 0.896797153024911, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 5.240857796842846e-06, | |
| "loss": 0.6145, | |
| "step": 3024 | |
| }, | |
| { | |
| "epoch": 0.8976868327402135, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 5.151668304058132e-06, | |
| "loss": 0.6146, | |
| "step": 3027 | |
| }, | |
| { | |
| "epoch": 0.8985765124555161, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 5.0632241887213275e-06, | |
| "loss": 0.6113, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.8994661921708185, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 4.975526145881515e-06, | |
| "loss": 0.5878, | |
| "step": 3033 | |
| }, | |
| { | |
| "epoch": 0.900355871886121, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.888574864724715e-06, | |
| "loss": 0.6173, | |
| "step": 3036 | |
| }, | |
| { | |
| "epoch": 0.9012455516014235, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 4.8023710285683975e-06, | |
| "loss": 0.5758, | |
| "step": 3039 | |
| }, | |
| { | |
| "epoch": 0.902135231316726, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 4.716915314856196e-06, | |
| "loss": 0.5931, | |
| "step": 3042 | |
| }, | |
| { | |
| "epoch": 0.9030249110320284, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.6322083951524705e-06, | |
| "loss": 0.5857, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 0.9039145907473309, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 4.548250935137144e-06, | |
| "loss": 0.5957, | |
| "step": 3048 | |
| }, | |
| { | |
| "epoch": 0.9048042704626335, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.465043594600382e-06, | |
| "loss": 0.6066, | |
| "step": 3051 | |
| }, | |
| { | |
| "epoch": 0.905693950177936, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 4.382587027437435e-06, | |
| "loss": 0.5993, | |
| "step": 3054 | |
| }, | |
| { | |
| "epoch": 0.9065836298932385, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 4.300881881643537e-06, | |
| "loss": 0.5769, | |
| "step": 3057 | |
| }, | |
| { | |
| "epoch": 0.9074733096085409, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 4.219928799308759e-06, | |
| "loss": 0.5997, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.9083629893238434, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 4.139728416613031e-06, | |
| "loss": 0.598, | |
| "step": 3063 | |
| }, | |
| { | |
| "epoch": 0.9092526690391459, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 4.0602813638210165e-06, | |
| "loss": 0.5925, | |
| "step": 3066 | |
| }, | |
| { | |
| "epoch": 0.9101423487544484, | |
| "grad_norm": 0.275390625, | |
| "learning_rate": 3.981588265277337e-06, | |
| "loss": 0.6136, | |
| "step": 3069 | |
| }, | |
| { | |
| "epoch": 0.9110320284697508, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 3.903649739401494e-06, | |
| "loss": 0.5873, | |
| "step": 3072 | |
| }, | |
| { | |
| "epoch": 0.9119217081850534, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 3.826466398683126e-06, | |
| "loss": 0.6177, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.9128113879003559, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 3.750038849677162e-06, | |
| "loss": 0.5871, | |
| "step": 3078 | |
| }, | |
| { | |
| "epoch": 0.9137010676156584, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.6743676929989924e-06, | |
| "loss": 0.5864, | |
| "step": 3081 | |
| }, | |
| { | |
| "epoch": 0.9145907473309609, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.5994535233198846e-06, | |
| "loss": 0.5827, | |
| "step": 3084 | |
| }, | |
| { | |
| "epoch": 0.9154804270462633, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.525296929362165e-06, | |
| "loss": 0.6044, | |
| "step": 3087 | |
| }, | |
| { | |
| "epoch": 0.9163701067615658, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 3.45189849389469e-06, | |
| "loss": 0.6116, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.9172597864768683, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 3.3792587937282128e-06, | |
| "loss": 0.5705, | |
| "step": 3093 | |
| }, | |
| { | |
| "epoch": 0.9181494661921709, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 3.30737839971087e-06, | |
| "loss": 0.6156, | |
| "step": 3096 | |
| }, | |
| { | |
| "epoch": 0.9190391459074733, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 3.236257876723725e-06, | |
| "loss": 0.5991, | |
| "step": 3099 | |
| }, | |
| { | |
| "epoch": 0.9199288256227758, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.165897783676275e-06, | |
| "loss": 0.5901, | |
| "step": 3102 | |
| }, | |
| { | |
| "epoch": 0.9208185053380783, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 3.0962986735020738e-06, | |
| "loss": 0.6183, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 0.9217081850533808, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.027461093154449e-06, | |
| "loss": 0.5892, | |
| "step": 3108 | |
| }, | |
| { | |
| "epoch": 0.9225978647686833, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.959385583602081e-06, | |
| "loss": 0.6269, | |
| "step": 3111 | |
| }, | |
| { | |
| "epoch": 0.9234875444839857, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.8920726798248643e-06, | |
| "loss": 0.5946, | |
| "step": 3114 | |
| }, | |
| { | |
| "epoch": 0.9243772241992882, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.8255229108096527e-06, | |
| "loss": 0.6192, | |
| "step": 3117 | |
| }, | |
| { | |
| "epoch": 0.9252669039145908, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.7597367995461086e-06, | |
| "loss": 0.6153, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.9261565836298933, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.694714863022585e-06, | |
| "loss": 0.5831, | |
| "step": 3123 | |
| }, | |
| { | |
| "epoch": 0.9270462633451957, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 2.6304576122221035e-06, | |
| "loss": 0.5898, | |
| "step": 3126 | |
| }, | |
| { | |
| "epoch": 0.9279359430604982, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 2.566965552118272e-06, | |
| "loss": 0.6098, | |
| "step": 3129 | |
| }, | |
| { | |
| "epoch": 0.9288256227758007, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 2.504239181671353e-06, | |
| "loss": 0.5932, | |
| "step": 3132 | |
| }, | |
| { | |
| "epoch": 0.9297153024911032, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.4422789938243763e-06, | |
| "loss": 0.5877, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 0.9306049822064056, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 2.381085475499201e-06, | |
| "loss": 0.5755, | |
| "step": 3138 | |
| }, | |
| { | |
| "epoch": 0.9314946619217082, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 2.3206591075927376e-06, | |
| "loss": 0.5875, | |
| "step": 3141 | |
| }, | |
| { | |
| "epoch": 0.9323843416370107, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 2.2610003649731092e-06, | |
| "loss": 0.6113, | |
| "step": 3144 | |
| }, | |
| { | |
| "epoch": 0.9332740213523132, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.2021097164760085e-06, | |
| "loss": 0.6035, | |
| "step": 3147 | |
| }, | |
| { | |
| "epoch": 0.9341637010676157, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 2.143987624900945e-06, | |
| "loss": 0.5813, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9350533807829181, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.0866345470076044e-06, | |
| "loss": 0.589, | |
| "step": 3153 | |
| }, | |
| { | |
| "epoch": 0.9359430604982206, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 2.0300509335123283e-06, | |
| "loss": 0.5971, | |
| "step": 3156 | |
| }, | |
| { | |
| "epoch": 0.9368327402135231, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.974237229084497e-06, | |
| "loss": 0.5808, | |
| "step": 3159 | |
| }, | |
| { | |
| "epoch": 0.9377224199288257, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.9191938723430615e-06, | |
| "loss": 0.6167, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 0.9386120996441281, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.8649212958531282e-06, | |
| "loss": 0.6088, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 0.9395017793594306, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.8114199261224928e-06, | |
| "loss": 0.5884, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 0.9403914590747331, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.7586901835983437e-06, | |
| "loss": 0.6122, | |
| "step": 3171 | |
| }, | |
| { | |
| "epoch": 0.9412811387900356, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 1.7067324826639419e-06, | |
| "loss": 0.6036, | |
| "step": 3174 | |
| }, | |
| { | |
| "epoch": 0.9421708185053381, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.655547231635368e-06, | |
| "loss": 0.598, | |
| "step": 3177 | |
| }, | |
| { | |
| "epoch": 0.9430604982206405, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.6051348327583037e-06, | |
| "loss": 0.6078, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.943950177935943, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 1.5554956822048661e-06, | |
| "loss": 0.5955, | |
| "step": 3183 | |
| }, | |
| { | |
| "epoch": 0.9448398576512456, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 1.5066301700705331e-06, | |
| "loss": 0.589, | |
| "step": 3186 | |
| }, | |
| { | |
| "epoch": 0.9457295373665481, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.4585386803710021e-06, | |
| "loss": 0.6035, | |
| "step": 3189 | |
| }, | |
| { | |
| "epoch": 0.9466192170818505, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.411221591039269e-06, | |
| "loss": 0.6396, | |
| "step": 3192 | |
| }, | |
| { | |
| "epoch": 0.947508896797153, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.3646792739225533e-06, | |
| "loss": 0.577, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 0.9483985765124555, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 1.3189120947794897e-06, | |
| "loss": 0.5983, | |
| "step": 3198 | |
| }, | |
| { | |
| "epoch": 0.949288256227758, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.273920413277152e-06, | |
| "loss": 0.6093, | |
| "step": 3201 | |
| }, | |
| { | |
| "epoch": 0.9501779359430605, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.2297045829882892e-06, | |
| "loss": 0.5966, | |
| "step": 3204 | |
| }, | |
| { | |
| "epoch": 0.951067615658363, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 1.186264951388516e-06, | |
| "loss": 0.6228, | |
| "step": 3207 | |
| }, | |
| { | |
| "epoch": 0.9519572953736655, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 1.1436018598535937e-06, | |
| "loss": 0.6083, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.952846975088968, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 1.1017156436567532e-06, | |
| "loss": 0.5806, | |
| "step": 3213 | |
| }, | |
| { | |
| "epoch": 0.9537366548042705, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 1.0606066319660435e-06, | |
| "loss": 0.579, | |
| "step": 3216 | |
| }, | |
| { | |
| "epoch": 0.9546263345195729, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 1.020275147841765e-06, | |
| "loss": 0.6053, | |
| "step": 3219 | |
| }, | |
| { | |
| "epoch": 0.9555160142348754, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 9.807215082339394e-07, | |
| "loss": 0.595, | |
| "step": 3222 | |
| }, | |
| { | |
| "epoch": 0.9564056939501779, | |
| "grad_norm": 0.25, | |
| "learning_rate": 9.41946023979745e-07, | |
| "loss": 0.5857, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.9572953736654805, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 9.039489998011852e-07, | |
| "loss": 0.6189, | |
| "step": 3228 | |
| }, | |
| { | |
| "epoch": 0.958185053380783, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.66730734302601e-07, | |
| "loss": 0.5837, | |
| "step": 3231 | |
| }, | |
| { | |
| "epoch": 0.9590747330960854, | |
| "grad_norm": 0.25, | |
| "learning_rate": 8.302915199683737e-07, | |
| "loss": 0.5827, | |
| "step": 3234 | |
| }, | |
| { | |
| "epoch": 0.9599644128113879, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 7.94631643160626e-07, | |
| "loss": 0.6043, | |
| "step": 3237 | |
| }, | |
| { | |
| "epoch": 0.9608540925266904, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 7.597513841169468e-07, | |
| "loss": 0.5621, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.9617437722419929, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 7.256510169482034e-07, | |
| "loss": 0.5886, | |
| "step": 3243 | |
| }, | |
| { | |
| "epoch": 0.9626334519572953, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.923308096363879e-07, | |
| "loss": 0.6205, | |
| "step": 3246 | |
| }, | |
| { | |
| "epoch": 0.9635231316725978, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 6.597910240324967e-07, | |
| "loss": 0.6038, | |
| "step": 3249 | |
| }, | |
| { | |
| "epoch": 0.9644128113879004, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 6.280319158544989e-07, | |
| "loss": 0.6301, | |
| "step": 3252 | |
| }, | |
| { | |
| "epoch": 0.9653024911032029, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.970537346853156e-07, | |
| "loss": 0.6007, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 0.9661921708185054, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 5.668567239708323e-07, | |
| "loss": 0.5789, | |
| "step": 3258 | |
| }, | |
| { | |
| "epoch": 0.9670818505338078, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 5.374411210180341e-07, | |
| "loss": 0.5964, | |
| "step": 3261 | |
| }, | |
| { | |
| "epoch": 0.9679715302491103, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 5.088071569931185e-07, | |
| "loss": 0.5953, | |
| "step": 3264 | |
| }, | |
| { | |
| "epoch": 0.9688612099644128, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 4.809550569196519e-07, | |
| "loss": 0.5877, | |
| "step": 3267 | |
| }, | |
| { | |
| "epoch": 0.9697508896797153, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 4.5388503967683793e-07, | |
| "loss": 0.5923, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.9706405693950177, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 4.275973179977855e-07, | |
| "loss": 0.5958, | |
| "step": 3273 | |
| }, | |
| { | |
| "epoch": 0.9715302491103203, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 4.0209209846783224e-07, | |
| "loss": 0.5977, | |
| "step": 3276 | |
| }, | |
| { | |
| "epoch": 0.9724199288256228, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 3.773695815229239e-07, | |
| "loss": 0.592, | |
| "step": 3279 | |
| }, | |
| { | |
| "epoch": 0.9733096085409253, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 3.534299614480596e-07, | |
| "loss": 0.5702, | |
| "step": 3282 | |
| }, | |
| { | |
| "epoch": 0.9741992882562278, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 3.3027342637572676e-07, | |
| "loss": 0.5893, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 0.9750889679715302, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.079001582844354e-07, | |
| "loss": 0.6177, | |
| "step": 3288 | |
| }, | |
| { | |
| "epoch": 0.9759786476868327, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 2.8631033299730825e-07, | |
| "loss": 0.6178, | |
| "step": 3291 | |
| }, | |
| { | |
| "epoch": 0.9768683274021353, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 2.655041201806707e-07, | |
| "loss": 0.5924, | |
| "step": 3294 | |
| }, | |
| { | |
| "epoch": 0.9777580071174378, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 2.454816833427631e-07, | |
| "loss": 0.6021, | |
| "step": 3297 | |
| }, | |
| { | |
| "epoch": 0.9786476868327402, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 2.2624317983239718e-07, | |
| "loss": 0.6131, | |
| "step": 3300 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 3372, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.1603485806523056e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |