| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 1480, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016891891891891893, | |
| "grad_norm": 6.5734802912605454, | |
| "learning_rate": 3.3783783783783788e-06, | |
| "loss": 1.3273, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.033783783783783786, | |
| "grad_norm": 6.1632920104999585, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 1.184, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05067567567567568, | |
| "grad_norm": 3.9191468793106843, | |
| "learning_rate": 1.0135135135135136e-05, | |
| "loss": 1.0141, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06756756756756757, | |
| "grad_norm": 6.836338097725109, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 0.9461, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08445945945945946, | |
| "grad_norm": 6.14454491803722, | |
| "learning_rate": 1.6891891891891892e-05, | |
| "loss": 0.9414, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10135135135135136, | |
| "grad_norm": 3.011838488661273, | |
| "learning_rate": 2.0270270270270273e-05, | |
| "loss": 0.9043, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11824324324324324, | |
| "grad_norm": 3.080629137119366, | |
| "learning_rate": 2.364864864864865e-05, | |
| "loss": 0.9097, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13513513513513514, | |
| "grad_norm": 2.8246920844764523, | |
| "learning_rate": 2.702702702702703e-05, | |
| "loss": 0.9112, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15202702702702703, | |
| "grad_norm": 2.8003645624320987, | |
| "learning_rate": 3.0405405405405407e-05, | |
| "loss": 0.9513, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16891891891891891, | |
| "grad_norm": 2.6192832071723293, | |
| "learning_rate": 3.3783783783783784e-05, | |
| "loss": 0.9393, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1858108108108108, | |
| "grad_norm": 2.6347662580290607, | |
| "learning_rate": 3.7162162162162165e-05, | |
| "loss": 0.9217, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.20270270270270271, | |
| "grad_norm": 2.6944921391744545, | |
| "learning_rate": 4.0540540540540545e-05, | |
| "loss": 0.9496, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2195945945945946, | |
| "grad_norm": 2.423675002168339, | |
| "learning_rate": 4.391891891891892e-05, | |
| "loss": 0.9423, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.23648648648648649, | |
| "grad_norm": 2.6803364488285326, | |
| "learning_rate": 4.72972972972973e-05, | |
| "loss": 0.9677, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2533783783783784, | |
| "grad_norm": 2.2781274624283685, | |
| "learning_rate": 4.999994383297182e-05, | |
| "loss": 0.966, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 2.4018446311363544, | |
| "learning_rate": 4.9997978016429605e-05, | |
| "loss": 0.9715, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.28716216216216217, | |
| "grad_norm": 2.3728527784166036, | |
| "learning_rate": 4.9993204128893056e-05, | |
| "loss": 0.977, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.30405405405405406, | |
| "grad_norm": 2.047106826030879, | |
| "learning_rate": 4.9985622766211935e-05, | |
| "loss": 0.9353, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.32094594594594594, | |
| "grad_norm": 2.1267608330495125, | |
| "learning_rate": 4.997523487464928e-05, | |
| "loss": 0.9432, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.33783783783783783, | |
| "grad_norm": 2.1225078532563537, | |
| "learning_rate": 4.996204175076325e-05, | |
| "loss": 0.9651, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3547297297297297, | |
| "grad_norm": 2.0337324708472906, | |
| "learning_rate": 4.99460450412453e-05, | |
| "loss": 0.9648, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3716216216216216, | |
| "grad_norm": 1.916581074879566, | |
| "learning_rate": 4.99272467427147e-05, | |
| "loss": 0.9675, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3885135135135135, | |
| "grad_norm": 2.121919880463027, | |
| "learning_rate": 4.990564920146926e-05, | |
| "loss": 0.9856, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.40540540540540543, | |
| "grad_norm": 2.214799458005234, | |
| "learning_rate": 4.9881255113192526e-05, | |
| "loss": 0.929, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4222972972972973, | |
| "grad_norm": 2.163453291125734, | |
| "learning_rate": 4.985406752261731e-05, | |
| "loss": 0.9473, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4391891891891892, | |
| "grad_norm": 1.9198145369029436, | |
| "learning_rate": 4.982408982314565e-05, | |
| "loss": 0.9678, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4560810810810811, | |
| "grad_norm": 2.0248877076532077, | |
| "learning_rate": 4.9791325756425264e-05, | |
| "loss": 0.9328, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.47297297297297297, | |
| "grad_norm": 1.807231483694689, | |
| "learning_rate": 4.975577941188258e-05, | |
| "loss": 0.9373, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48986486486486486, | |
| "grad_norm": 1.8706434071195257, | |
| "learning_rate": 4.971745522621225e-05, | |
| "loss": 0.9626, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5067567567567568, | |
| "grad_norm": 1.7945270289752226, | |
| "learning_rate": 4.967635798282344e-05, | |
| "loss": 0.9705, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5236486486486487, | |
| "grad_norm": 1.8616405046783886, | |
| "learning_rate": 4.963249281124278e-05, | |
| "loss": 0.9072, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 1.666683134269227, | |
| "learning_rate": 4.958586518647407e-05, | |
| "loss": 0.898, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5574324324324325, | |
| "grad_norm": 1.594107803207033, | |
| "learning_rate": 4.953648092831505e-05, | |
| "loss": 0.9251, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5743243243243243, | |
| "grad_norm": 1.8741745620159964, | |
| "learning_rate": 4.9484346200630855e-05, | |
| "loss": 0.9687, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5912162162162162, | |
| "grad_norm": 1.905205634059141, | |
| "learning_rate": 4.9429467510584794e-05, | |
| "loss": 0.8999, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6081081081081081, | |
| "grad_norm": 1.9962261429841193, | |
| "learning_rate": 4.937185170782607e-05, | |
| "loss": 0.9368, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 2.1664869693090165, | |
| "learning_rate": 4.931150598363494e-05, | |
| "loss": 0.9061, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6418918918918919, | |
| "grad_norm": 1.6870468109861425, | |
| "learning_rate": 4.9248437870025035e-05, | |
| "loss": 0.8989, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6587837837837838, | |
| "grad_norm": 1.6966567361912948, | |
| "learning_rate": 4.9182655238803365e-05, | |
| "loss": 0.8989, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 1.692847679116018, | |
| "learning_rate": 4.911416630058772e-05, | |
| "loss": 0.8889, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6925675675675675, | |
| "grad_norm": 1.666285584632242, | |
| "learning_rate": 4.9042979603781884e-05, | |
| "loss": 0.9038, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7094594594594594, | |
| "grad_norm": 1.7022768095416743, | |
| "learning_rate": 4.896910403350873e-05, | |
| "loss": 0.9247, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7263513513513513, | |
| "grad_norm": 1.6583467639807945, | |
| "learning_rate": 4.8892548810501146e-05, | |
| "loss": 0.9042, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7432432432432432, | |
| "grad_norm": 1.5684890837209118, | |
| "learning_rate": 4.88133234899512e-05, | |
| "loss": 0.8875, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7601351351351351, | |
| "grad_norm": 1.7067630214310312, | |
| "learning_rate": 4.873143796031752e-05, | |
| "loss": 0.8797, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.777027027027027, | |
| "grad_norm": 1.5501274222949621, | |
| "learning_rate": 4.864690244209105e-05, | |
| "loss": 0.8903, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.793918918918919, | |
| "grad_norm": 1.6739236540233466, | |
| "learning_rate": 4.855972748651939e-05, | |
| "loss": 0.9036, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 1.5251028899529802, | |
| "learning_rate": 4.8469923974289874e-05, | |
| "loss": 0.876, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8277027027027027, | |
| "grad_norm": 1.6416347283581953, | |
| "learning_rate": 4.837750311417146e-05, | |
| "loss": 0.8777, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8445945945945946, | |
| "grad_norm": 1.7148770313619208, | |
| "learning_rate": 4.828247644161577e-05, | |
| "loss": 0.9305, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8614864864864865, | |
| "grad_norm": 1.6917833006591425, | |
| "learning_rate": 4.8184855817317226e-05, | |
| "loss": 0.8902, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8783783783783784, | |
| "grad_norm": 1.4551269904823192, | |
| "learning_rate": 4.808465342573274e-05, | |
| "loss": 0.866, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8952702702702703, | |
| "grad_norm": 1.6585601416449522, | |
| "learning_rate": 4.7981881773560886e-05, | |
| "loss": 0.896, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9121621621621622, | |
| "grad_norm": 1.5413455024296299, | |
| "learning_rate": 4.787655368818087e-05, | |
| "loss": 0.8684, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9290540540540541, | |
| "grad_norm": 1.4770164865772377, | |
| "learning_rate": 4.77686823160515e-05, | |
| "loss": 0.8704, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9459459459459459, | |
| "grad_norm": 1.642857550796514, | |
| "learning_rate": 4.765828112107034e-05, | |
| "loss": 0.8749, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9628378378378378, | |
| "grad_norm": 2.028272824380857, | |
| "learning_rate": 4.75453638828932e-05, | |
| "loss": 0.8654, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9797297297297297, | |
| "grad_norm": 1.4679639397479571, | |
| "learning_rate": 4.742994469521421e-05, | |
| "loss": 0.8601, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9966216216216216, | |
| "grad_norm": 1.4327374675646831, | |
| "learning_rate": 4.7312037964006806e-05, | |
| "loss": 0.8564, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.0135135135135136, | |
| "grad_norm": 1.6103445933566929, | |
| "learning_rate": 4.719165840572557e-05, | |
| "loss": 0.6356, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0304054054054055, | |
| "grad_norm": 1.2609614774524984, | |
| "learning_rate": 4.7068821045469464e-05, | |
| "loss": 0.5569, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0472972972972974, | |
| "grad_norm": 1.6175496296524783, | |
| "learning_rate": 4.694354121510644e-05, | |
| "loss": 0.5491, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0641891891891893, | |
| "grad_norm": 1.426861182551709, | |
| "learning_rate": 4.681583455135985e-05, | |
| "loss": 0.5354, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0810810810810811, | |
| "grad_norm": 1.3646672710235592, | |
| "learning_rate": 4.668571699385668e-05, | |
| "loss": 0.5427, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.097972972972973, | |
| "grad_norm": 1.5346272239009828, | |
| "learning_rate": 4.655320478313817e-05, | |
| "loss": 0.5359, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.114864864864865, | |
| "grad_norm": 1.3759835475947566, | |
| "learning_rate": 4.641831445863265e-05, | |
| "loss": 0.5638, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1317567567567568, | |
| "grad_norm": 1.4909239756270425, | |
| "learning_rate": 4.628106285659124e-05, | |
| "loss": 0.5741, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1486486486486487, | |
| "grad_norm": 1.7461008672361147, | |
| "learning_rate": 4.614146710798645e-05, | |
| "loss": 0.5461, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1655405405405406, | |
| "grad_norm": 1.5678203614090302, | |
| "learning_rate": 4.599954463637394e-05, | |
| "loss": 0.5898, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1824324324324325, | |
| "grad_norm": 1.3911715983555075, | |
| "learning_rate": 4.585531315571788e-05, | |
| "loss": 0.5478, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1993243243243243, | |
| "grad_norm": 1.521721643927216, | |
| "learning_rate": 4.570879066817991e-05, | |
| "loss": 0.5528, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.2162162162162162, | |
| "grad_norm": 1.4667462390466979, | |
| "learning_rate": 4.555999546187229e-05, | |
| "loss": 0.5754, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.2331081081081081, | |
| "grad_norm": 1.5550072973306515, | |
| "learning_rate": 4.5408946108575215e-05, | |
| "loss": 0.5708, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.3431454538729617, | |
| "learning_rate": 4.5255661461418854e-05, | |
| "loss": 0.5577, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2668918918918919, | |
| "grad_norm": 1.4814401713234768, | |
| "learning_rate": 4.510016065253016e-05, | |
| "loss": 0.5643, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2837837837837838, | |
| "grad_norm": 1.3784794077594908, | |
| "learning_rate": 4.4942463090644896e-05, | |
| "loss": 0.5925, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.3006756756756757, | |
| "grad_norm": 1.3720445672478283, | |
| "learning_rate": 4.478258845868522e-05, | |
| "loss": 0.5807, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.3175675675675675, | |
| "grad_norm": 1.4533572289367873, | |
| "learning_rate": 4.462055671130289e-05, | |
| "loss": 0.5796, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3344594594594594, | |
| "grad_norm": 1.5146363381910304, | |
| "learning_rate": 4.445638807238865e-05, | |
| "loss": 0.5804, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 1.5458264427252129, | |
| "learning_rate": 4.4290103032548094e-05, | |
| "loss": 0.5722, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3682432432432432, | |
| "grad_norm": 1.3173722679798987, | |
| "learning_rate": 4.412172234654399e-05, | |
| "loss": 0.5531, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.385135135135135, | |
| "grad_norm": 1.4883033189800094, | |
| "learning_rate": 4.395126703070589e-05, | |
| "loss": 0.5577, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.402027027027027, | |
| "grad_norm": 1.3909336605285059, | |
| "learning_rate": 4.3778758360306986e-05, | |
| "loss": 0.578, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.4189189189189189, | |
| "grad_norm": 1.3005774948488031, | |
| "learning_rate": 4.360421786690862e-05, | |
| "loss": 0.5583, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4358108108108107, | |
| "grad_norm": 1.4342849633761854, | |
| "learning_rate": 4.3427667335672815e-05, | |
| "loss": 0.5842, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4527027027027026, | |
| "grad_norm": 1.4154466456437151, | |
| "learning_rate": 4.324912880264326e-05, | |
| "loss": 0.5491, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4695945945945945, | |
| "grad_norm": 1.4449023037096216, | |
| "learning_rate": 4.30686245519948e-05, | |
| "loss": 0.5459, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.4864864864864864, | |
| "grad_norm": 1.2746647387428396, | |
| "learning_rate": 4.288617711325207e-05, | |
| "loss": 0.5819, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.5033783783783785, | |
| "grad_norm": 1.447467776474505, | |
| "learning_rate": 4.2701809258477575e-05, | |
| "loss": 0.5784, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.5202702702702702, | |
| "grad_norm": 1.2391202211909191, | |
| "learning_rate": 4.251554399942928e-05, | |
| "loss": 0.5701, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.5371621621621623, | |
| "grad_norm": 1.2268682468986518, | |
| "learning_rate": 4.2327404584688495e-05, | |
| "loss": 0.5995, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.554054054054054, | |
| "grad_norm": 1.3918773327038636, | |
| "learning_rate": 4.21374144967581e-05, | |
| "loss": 0.5644, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.570945945945946, | |
| "grad_norm": 1.4547926883844007, | |
| "learning_rate": 4.194559744913157e-05, | |
| "loss": 0.5642, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.5878378378378377, | |
| "grad_norm": 1.3596176232205965, | |
| "learning_rate": 4.1751977383333224e-05, | |
| "loss": 0.5591, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.6047297297297298, | |
| "grad_norm": 1.3130436197901627, | |
| "learning_rate": 4.1556578465929955e-05, | |
| "loss": 0.571, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.6216216216216215, | |
| "grad_norm": 1.402701522056968, | |
| "learning_rate": 4.1359425085514906e-05, | |
| "loss": 0.5927, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.6385135135135136, | |
| "grad_norm": 1.2781385628050617, | |
| "learning_rate": 4.1160541849663404e-05, | |
| "loss": 0.5745, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.6554054054054053, | |
| "grad_norm": 1.3121044373462736, | |
| "learning_rate": 4.095995358186162e-05, | |
| "loss": 0.5734, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6722972972972974, | |
| "grad_norm": 1.3358239819906057, | |
| "learning_rate": 4.0757685318408186e-05, | |
| "loss": 0.5553, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.689189189189189, | |
| "grad_norm": 1.3109247815754774, | |
| "learning_rate": 4.055376230528936e-05, | |
| "loss": 0.5621, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.7060810810810811, | |
| "grad_norm": 1.3687138517149326, | |
| "learning_rate": 4.034820999502794e-05, | |
| "loss": 0.571, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.722972972972973, | |
| "grad_norm": 1.2988387849104504, | |
| "learning_rate": 4.0141054043506406e-05, | |
| "loss": 0.5741, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.739864864864865, | |
| "grad_norm": 1.3411202779787594, | |
| "learning_rate": 3.993232030676473e-05, | |
| "loss": 0.5685, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.7567567567567568, | |
| "grad_norm": 1.2949335901614112, | |
| "learning_rate": 3.972203483777315e-05, | |
| "loss": 0.558, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7736486486486487, | |
| "grad_norm": 1.3721749127594483, | |
| "learning_rate": 3.95102238831804e-05, | |
| "loss": 0.5915, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.7905405405405406, | |
| "grad_norm": 1.350082387281707, | |
| "learning_rate": 3.929691388003772e-05, | |
| "loss": 0.5685, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.8074324324324325, | |
| "grad_norm": 1.3245276027947501, | |
| "learning_rate": 3.908213145249916e-05, | |
| "loss": 0.5691, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.8243243243243243, | |
| "grad_norm": 1.4137310799251641, | |
| "learning_rate": 3.886590340849852e-05, | |
| "loss": 0.566, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.8412162162162162, | |
| "grad_norm": 1.356555489682696, | |
| "learning_rate": 3.864825673640326e-05, | |
| "loss": 0.551, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.8581081081081081, | |
| "grad_norm": 1.2815437787372281, | |
| "learning_rate": 3.842921860164607e-05, | |
| "loss": 0.5655, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 1.3576111263757158, | |
| "learning_rate": 3.8208816343334156e-05, | |
| "loss": 0.5497, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.8918918918918919, | |
| "grad_norm": 1.2750346506851613, | |
| "learning_rate": 3.798707747083694e-05, | |
| "loss": 0.5746, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.9087837837837838, | |
| "grad_norm": 1.317301287962323, | |
| "learning_rate": 3.776402966035251e-05, | |
| "loss": 0.584, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.9256756756756757, | |
| "grad_norm": 1.2204583503286293, | |
| "learning_rate": 3.753970075145322e-05, | |
| "loss": 0.5518, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.9425675675675675, | |
| "grad_norm": 1.2700469074124594, | |
| "learning_rate": 3.731411874361094e-05, | |
| "loss": 0.5466, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.9594594594594594, | |
| "grad_norm": 1.2657661298703708, | |
| "learning_rate": 3.7087311792702265e-05, | |
| "loss": 0.5659, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9763513513513513, | |
| "grad_norm": 1.2536149686511529, | |
| "learning_rate": 3.685930820749433e-05, | |
| "loss": 0.5555, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.9932432432432432, | |
| "grad_norm": 1.2809671695628992, | |
| "learning_rate": 3.663013644611139e-05, | |
| "loss": 0.5563, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.010135135135135, | |
| "grad_norm": 1.103796522043604, | |
| "learning_rate": 3.639982511248289e-05, | |
| "loss": 0.3723, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.027027027027027, | |
| "grad_norm": 1.2569696258546523, | |
| "learning_rate": 3.616840295277328e-05, | |
| "loss": 0.2453, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.043918918918919, | |
| "grad_norm": 1.202522926688232, | |
| "learning_rate": 3.593589885179405e-05, | |
| "loss": 0.2582, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.060810810810811, | |
| "grad_norm": 1.141634620637362, | |
| "learning_rate": 3.5702341829398525e-05, | |
| "loss": 0.25, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0777027027027026, | |
| "grad_norm": 1.2434462048859511, | |
| "learning_rate": 3.5467761036859736e-05, | |
| "loss": 0.2524, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.0945945945945947, | |
| "grad_norm": 1.2503774124889013, | |
| "learning_rate": 3.523218575323198e-05, | |
| "loss": 0.2426, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.1114864864864864, | |
| "grad_norm": 1.0984336540308752, | |
| "learning_rate": 3.499564538169629e-05, | |
| "loss": 0.2442, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.1283783783783785, | |
| "grad_norm": 1.3274920050931374, | |
| "learning_rate": 3.475816944589058e-05, | |
| "loss": 0.2426, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.14527027027027, | |
| "grad_norm": 1.2086016497404501, | |
| "learning_rate": 3.451978758622458e-05, | |
| "loss": 0.2549, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.1621621621621623, | |
| "grad_norm": 1.1812293557905968, | |
| "learning_rate": 3.4280529556180404e-05, | |
| "loss": 0.2383, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.179054054054054, | |
| "grad_norm": 1.4704829661444183, | |
| "learning_rate": 3.4040425218598755e-05, | |
| "loss": 0.2441, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.195945945945946, | |
| "grad_norm": 1.2763011019642492, | |
| "learning_rate": 3.379950454195172e-05, | |
| "loss": 0.252, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.2128378378378377, | |
| "grad_norm": 1.2711259414467864, | |
| "learning_rate": 3.355779759660223e-05, | |
| "loss": 0.2482, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.22972972972973, | |
| "grad_norm": 1.2196154702915583, | |
| "learning_rate": 3.331533455105084e-05, | |
| "loss": 0.2422, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.2466216216216215, | |
| "grad_norm": 1.0928327721884867, | |
| "learning_rate": 3.307214566817027e-05, | |
| "loss": 0.2493, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.2635135135135136, | |
| "grad_norm": 1.1577331879228163, | |
| "learning_rate": 3.2828261301428206e-05, | |
| "loss": 0.2547, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.2804054054054053, | |
| "grad_norm": 1.142585348862718, | |
| "learning_rate": 3.2583711891098665e-05, | |
| "loss": 0.2545, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.2972972972972974, | |
| "grad_norm": 1.3931724258065883, | |
| "learning_rate": 3.23385279604627e-05, | |
| "loss": 0.2555, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.314189189189189, | |
| "grad_norm": 1.12997531420638, | |
| "learning_rate": 3.209274011199861e-05, | |
| "loss": 0.2503, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.331081081081081, | |
| "grad_norm": 1.1911209479824547, | |
| "learning_rate": 3.18463790235623e-05, | |
| "loss": 0.252, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.347972972972973, | |
| "grad_norm": 1.2118767972123496, | |
| "learning_rate": 3.159947544455828e-05, | |
| "loss": 0.2489, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.364864864864865, | |
| "grad_norm": 1.1732170281780285, | |
| "learning_rate": 3.135206019210167e-05, | |
| "loss": 0.2452, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3817567567567566, | |
| "grad_norm": 1.171003404202182, | |
| "learning_rate": 3.110416414717181e-05, | |
| "loss": 0.2565, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.3986486486486487, | |
| "grad_norm": 1.087800299348181, | |
| "learning_rate": 3.085581825075782e-05, | |
| "loss": 0.2428, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.4155405405405403, | |
| "grad_norm": 1.1798352511221857, | |
| "learning_rate": 3.060705349999677e-05, | |
| "loss": 0.2597, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.4324324324324325, | |
| "grad_norm": 1.1164619974303382, | |
| "learning_rate": 3.0357900944304774e-05, | |
| "loss": 0.2558, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.4493243243243246, | |
| "grad_norm": 1.1783745618133863, | |
| "learning_rate": 3.0108391681501564e-05, | |
| "loss": 0.2502, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.4662162162162162, | |
| "grad_norm": 1.2059356558094934, | |
| "learning_rate": 2.9858556853929048e-05, | |
| "loss": 0.2549, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.483108108108108, | |
| "grad_norm": 1.1084513675225853, | |
| "learning_rate": 2.96084276445643e-05, | |
| "loss": 0.2623, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.1268704196768717, | |
| "learning_rate": 2.9358035273127483e-05, | |
| "loss": 0.2445, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.516891891891892, | |
| "grad_norm": 1.2073238153053072, | |
| "learning_rate": 2.910741099218514e-05, | |
| "loss": 0.2523, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.5337837837837838, | |
| "grad_norm": 1.1875133895260694, | |
| "learning_rate": 2.8856586083249487e-05, | |
| "loss": 0.2587, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.5506756756756754, | |
| "grad_norm": 1.1828390695293125, | |
| "learning_rate": 2.860559185287397e-05, | |
| "loss": 0.2555, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.5675675675675675, | |
| "grad_norm": 1.1686470252895278, | |
| "learning_rate": 2.83544596287458e-05, | |
| "loss": 0.246, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5844594594594597, | |
| "grad_norm": 1.2445669104213888, | |
| "learning_rate": 2.8103220755775776e-05, | |
| "loss": 0.2536, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.6013513513513513, | |
| "grad_norm": 1.1612663427931047, | |
| "learning_rate": 2.785190659218604e-05, | |
| "loss": 0.2511, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.618243243243243, | |
| "grad_norm": 1.169020666404516, | |
| "learning_rate": 2.760054850559603e-05, | |
| "loss": 0.2464, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.635135135135135, | |
| "grad_norm": 1.1211602172208788, | |
| "learning_rate": 2.7349177869107462e-05, | |
| "loss": 0.2425, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.652027027027027, | |
| "grad_norm": 1.1295937771546627, | |
| "learning_rate": 2.709782605738842e-05, | |
| "loss": 0.251, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.668918918918919, | |
| "grad_norm": 1.1985208404439514, | |
| "learning_rate": 2.684652444275741e-05, | |
| "loss": 0.2511, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.685810810810811, | |
| "grad_norm": 1.1712043419573621, | |
| "learning_rate": 2.6595304391267605e-05, | |
| "loss": 0.2409, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 1.1600530170137202, | |
| "learning_rate": 2.634419725879193e-05, | |
| "loss": 0.2461, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.7195945945945947, | |
| "grad_norm": 1.0980080565697286, | |
| "learning_rate": 2.60932343871094e-05, | |
| "loss": 0.2351, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.7364864864864864, | |
| "grad_norm": 1.204458523946955, | |
| "learning_rate": 2.58424470999932e-05, | |
| "loss": 0.2415, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.7533783783783785, | |
| "grad_norm": 1.2137349551053236, | |
| "learning_rate": 2.5591866699301055e-05, | |
| "loss": 0.2411, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.77027027027027, | |
| "grad_norm": 1.2375233171662754, | |
| "learning_rate": 2.534152446106825e-05, | |
| "loss": 0.2429, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7871621621621623, | |
| "grad_norm": 1.1646244068445126, | |
| "learning_rate": 2.5091451631604033e-05, | |
| "loss": 0.2359, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.804054054054054, | |
| "grad_norm": 1.1619752616590404, | |
| "learning_rate": 2.4841679423591523e-05, | |
| "loss": 0.2368, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.820945945945946, | |
| "grad_norm": 1.1047288053357993, | |
| "learning_rate": 2.4592239012191977e-05, | |
| "loss": 0.2391, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.8378378378378377, | |
| "grad_norm": 1.074588438071139, | |
| "learning_rate": 2.4343161531153647e-05, | |
| "loss": 0.234, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.85472972972973, | |
| "grad_norm": 1.1971487515181058, | |
| "learning_rate": 2.4094478068925884e-05, | |
| "loss": 0.2459, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.8716216216216215, | |
| "grad_norm": 1.18752516936673, | |
| "learning_rate": 2.3846219664778824e-05, | |
| "loss": 0.2337, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.8885135135135136, | |
| "grad_norm": 1.0648986336743749, | |
| "learning_rate": 2.3598417304929226e-05, | |
| "loss": 0.2385, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.9054054054054053, | |
| "grad_norm": 1.2273436401598625, | |
| "learning_rate": 2.3351101918672985e-05, | |
| "loss": 0.2307, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.9222972972972974, | |
| "grad_norm": 1.1562044904763753, | |
| "learning_rate": 2.3104304374524704e-05, | |
| "loss": 0.2313, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.939189189189189, | |
| "grad_norm": 1.1200371062184964, | |
| "learning_rate": 2.2858055476364822e-05, | |
| "loss": 0.2382, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.956081081081081, | |
| "grad_norm": 1.0863337573648397, | |
| "learning_rate": 2.2612385959594877e-05, | |
| "loss": 0.2357, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.972972972972973, | |
| "grad_norm": 1.1238376465153148, | |
| "learning_rate": 2.2367326487301317e-05, | |
| "loss": 0.2366, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.989864864864865, | |
| "grad_norm": 1.1679762876800548, | |
| "learning_rate": 2.2122907646428214e-05, | |
| "loss": 0.233, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.0067567567567566, | |
| "grad_norm": 0.7617860489835798, | |
| "learning_rate": 2.1879159943959686e-05, | |
| "loss": 0.1782, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.0236486486486487, | |
| "grad_norm": 0.7980232728533272, | |
| "learning_rate": 2.1636113803112097e-05, | |
| "loss": 0.0889, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.0405405405405403, | |
| "grad_norm": 1.0371057007455198, | |
| "learning_rate": 2.139379955953686e-05, | |
| "loss": 0.0855, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.0574324324324325, | |
| "grad_norm": 0.922694869510506, | |
| "learning_rate": 2.1152247457534065e-05, | |
| "loss": 0.082, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.074324324324324, | |
| "grad_norm": 0.8665469624615858, | |
| "learning_rate": 2.0911487646277623e-05, | |
| "loss": 0.0842, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.0912162162162162, | |
| "grad_norm": 0.8466103971895397, | |
| "learning_rate": 2.067155017605212e-05, | |
| "loss": 0.0818, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.108108108108108, | |
| "grad_norm": 0.8200266011697321, | |
| "learning_rate": 2.0432464994502203e-05, | |
| "loss": 0.0839, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 0.805011414682386, | |
| "learning_rate": 2.0194261942894628e-05, | |
| "loss": 0.0798, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.141891891891892, | |
| "grad_norm": 0.9572346016824955, | |
| "learning_rate": 1.995697075239365e-05, | |
| "loss": 0.0856, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.1587837837837838, | |
| "grad_norm": 0.8204526811751627, | |
| "learning_rate": 1.972062104035017e-05, | |
| "loss": 0.079, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.175675675675676, | |
| "grad_norm": 0.8185257809669213, | |
| "learning_rate": 1.9485242306605028e-05, | |
| "loss": 0.0865, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.1925675675675675, | |
| "grad_norm": 0.8796508468976082, | |
| "learning_rate": 1.9250863929807027e-05, | |
| "loss": 0.0786, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.2094594594594597, | |
| "grad_norm": 0.8756829385022001, | |
| "learning_rate": 1.9017515163746058e-05, | |
| "loss": 0.0836, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.2263513513513513, | |
| "grad_norm": 0.8157354792396803, | |
| "learning_rate": 1.878522513370177e-05, | |
| "loss": 0.0852, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.2432432432432434, | |
| "grad_norm": 0.8590600023617616, | |
| "learning_rate": 1.855402283280836e-05, | |
| "loss": 0.0835, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.260135135135135, | |
| "grad_norm": 0.8643752542390469, | |
| "learning_rate": 1.8323937118435786e-05, | |
| "loss": 0.0814, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.277027027027027, | |
| "grad_norm": 0.9225152379758252, | |
| "learning_rate": 1.8094996708587958e-05, | |
| "loss": 0.0861, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.293918918918919, | |
| "grad_norm": 0.9044834326870717, | |
| "learning_rate": 1.7867230178318334e-05, | |
| "loss": 0.0819, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.310810810810811, | |
| "grad_norm": 0.8920506863920826, | |
| "learning_rate": 1.7640665956163306e-05, | |
| "loss": 0.0809, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.3277027027027026, | |
| "grad_norm": 0.9035029361545068, | |
| "learning_rate": 1.7415332320593964e-05, | |
| "loss": 0.0782, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.3445945945945947, | |
| "grad_norm": 0.8458923715737863, | |
| "learning_rate": 1.719125739648648e-05, | |
| "loss": 0.0795, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.3614864864864864, | |
| "grad_norm": 0.8283053816212952, | |
| "learning_rate": 1.6968469151611766e-05, | |
| "loss": 0.0766, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "grad_norm": 0.8124298394489518, | |
| "learning_rate": 1.6746995393144668e-05, | |
| "loss": 0.0795, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.39527027027027, | |
| "grad_norm": 0.84974780220989, | |
| "learning_rate": 1.6526863764193228e-05, | |
| "loss": 0.0816, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 3.4121621621621623, | |
| "grad_norm": 0.9040056390996183, | |
| "learning_rate": 1.6308101740348433e-05, | |
| "loss": 0.0862, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.429054054054054, | |
| "grad_norm": 0.8503065695427434, | |
| "learning_rate": 1.6090736626254894e-05, | |
| "loss": 0.0821, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 3.445945945945946, | |
| "grad_norm": 0.922976948565284, | |
| "learning_rate": 1.5874795552202773e-05, | |
| "loss": 0.0767, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.4628378378378377, | |
| "grad_norm": 0.8359859623361058, | |
| "learning_rate": 1.5660305470741603e-05, | |
| "loss": 0.0822, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.47972972972973, | |
| "grad_norm": 0.7993044555598261, | |
| "learning_rate": 1.5447293153316163e-05, | |
| "loss": 0.0725, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.4966216216216215, | |
| "grad_norm": 0.8837194556202582, | |
| "learning_rate": 1.523578518692505e-05, | |
| "loss": 0.0803, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 3.5135135135135136, | |
| "grad_norm": 0.8671077384687282, | |
| "learning_rate": 1.5025807970802252e-05, | |
| "loss": 0.0759, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.5304054054054053, | |
| "grad_norm": 0.851094402030443, | |
| "learning_rate": 1.481738771312209e-05, | |
| "loss": 0.079, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 3.5472972972972974, | |
| "grad_norm": 0.8361299277027681, | |
| "learning_rate": 1.4610550427728103e-05, | |
| "loss": 0.076, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.564189189189189, | |
| "grad_norm": 0.8472817077547173, | |
| "learning_rate": 1.4405321930886161e-05, | |
| "loss": 0.0762, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 3.581081081081081, | |
| "grad_norm": 0.8496742173511822, | |
| "learning_rate": 1.4201727838062181e-05, | |
| "loss": 0.0777, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.597972972972973, | |
| "grad_norm": 0.9073909261434937, | |
| "learning_rate": 1.3999793560724966e-05, | |
| "loss": 0.0778, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 3.614864864864865, | |
| "grad_norm": 0.8148037450824118, | |
| "learning_rate": 1.3799544303174514e-05, | |
| "loss": 0.0753, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.631756756756757, | |
| "grad_norm": 0.792903387483005, | |
| "learning_rate": 1.3601005059396104e-05, | |
| "loss": 0.0743, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.6486486486486487, | |
| "grad_norm": 0.8138776105456261, | |
| "learning_rate": 1.3404200609940754e-05, | |
| "loss": 0.0771, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.6655405405405403, | |
| "grad_norm": 0.8412665781860845, | |
| "learning_rate": 1.3209155518832203e-05, | |
| "loss": 0.0748, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 3.6824324324324325, | |
| "grad_norm": 0.8443573738089697, | |
| "learning_rate": 1.3015894130500977e-05, | |
| "loss": 0.0729, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.6993243243243246, | |
| "grad_norm": 0.8610810221368033, | |
| "learning_rate": 1.2824440566745865e-05, | |
| "loss": 0.0744, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 3.7162162162162162, | |
| "grad_norm": 0.7968023292862391, | |
| "learning_rate": 1.2634818723723174e-05, | |
| "loss": 0.0752, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.733108108108108, | |
| "grad_norm": 0.7623771883673489, | |
| "learning_rate": 1.2447052268964122e-05, | |
| "loss": 0.0698, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.8911604401532162, | |
| "learning_rate": 1.2261164638420832e-05, | |
| "loss": 0.0728, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.766891891891892, | |
| "grad_norm": 0.8291626224492225, | |
| "learning_rate": 1.2077179033541139e-05, | |
| "loss": 0.0725, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 3.7837837837837838, | |
| "grad_norm": 0.7888246317862425, | |
| "learning_rate": 1.1895118418372734e-05, | |
| "loss": 0.0751, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.8006756756756754, | |
| "grad_norm": 0.7596045011769311, | |
| "learning_rate": 1.171500551669697e-05, | |
| "loss": 0.0707, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.8175675675675675, | |
| "grad_norm": 0.8261890861433314, | |
| "learning_rate": 1.1536862809192518e-05, | |
| "loss": 0.0699, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.8344594594594597, | |
| "grad_norm": 0.8157898272197499, | |
| "learning_rate": 1.1360712530629513e-05, | |
| "loss": 0.0739, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 3.8513513513513513, | |
| "grad_norm": 0.7953827120772039, | |
| "learning_rate": 1.1186576667094342e-05, | |
| "loss": 0.0727, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.868243243243243, | |
| "grad_norm": 0.8106925387796762, | |
| "learning_rate": 1.1014476953245423e-05, | |
| "loss": 0.0689, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 3.885135135135135, | |
| "grad_norm": 0.7915344112028949, | |
| "learning_rate": 1.0844434869600428e-05, | |
| "loss": 0.0696, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.902027027027027, | |
| "grad_norm": 0.8231236378626898, | |
| "learning_rate": 1.0676471639855235e-05, | |
| "loss": 0.0701, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 3.918918918918919, | |
| "grad_norm": 0.8574639286128795, | |
| "learning_rate": 1.0510608228234848e-05, | |
| "loss": 0.0678, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.935810810810811, | |
| "grad_norm": 0.7703198418021956, | |
| "learning_rate": 1.0346865336876809e-05, | |
| "loss": 0.0679, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 3.9527027027027026, | |
| "grad_norm": 0.8086225426905125, | |
| "learning_rate": 1.0185263403247256e-05, | |
| "loss": 0.0669, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.9695945945945947, | |
| "grad_norm": 0.8488002772643025, | |
| "learning_rate": 1.002582259759002e-05, | |
| "loss": 0.0684, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.9864864864864864, | |
| "grad_norm": 0.8980288101695698, | |
| "learning_rate": 9.868562820409103e-06, | |
| "loss": 0.0686, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.003378378378378, | |
| "grad_norm": 0.3986041630321839, | |
| "learning_rate": 9.713503699984825e-06, | |
| "loss": 0.0548, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.02027027027027, | |
| "grad_norm": 0.43726382965712174, | |
| "learning_rate": 9.560664589923895e-06, | |
| "loss": 0.0219, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.037162162162162, | |
| "grad_norm": 0.36710468007751074, | |
| "learning_rate": 9.410064566743841e-06, | |
| "loss": 0.0208, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.054054054054054, | |
| "grad_norm": 0.37287043834241884, | |
| "learning_rate": 9.261722427491953e-06, | |
| "loss": 0.0194, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.070945945945946, | |
| "grad_norm": 0.5533039626983974, | |
| "learning_rate": 9.115656687399149e-06, | |
| "loss": 0.02, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.087837837837838, | |
| "grad_norm": 0.5313826064857297, | |
| "learning_rate": 8.971885577569058e-06, | |
| "loss": 0.0177, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.10472972972973, | |
| "grad_norm": 0.44890915578901536, | |
| "learning_rate": 8.83042704270243e-06, | |
| "loss": 0.0189, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.121621621621622, | |
| "grad_norm": 0.5560917251637355, | |
| "learning_rate": 8.691298738857432e-06, | |
| "loss": 0.019, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.138513513513513, | |
| "grad_norm": 0.5436237010579721, | |
| "learning_rate": 8.554518031245934e-06, | |
| "loss": 0.0185, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.155405405405405, | |
| "grad_norm": 0.4740850367041631, | |
| "learning_rate": 8.420101992066028e-06, | |
| "loss": 0.0166, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.172297297297297, | |
| "grad_norm": 0.49052875884939057, | |
| "learning_rate": 8.288067398371214e-06, | |
| "loss": 0.0167, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 4.1891891891891895, | |
| "grad_norm": 0.43052256280300827, | |
| "learning_rate": 8.158430729976372e-06, | |
| "loss": 0.0186, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.206081081081081, | |
| "grad_norm": 0.4777607693530672, | |
| "learning_rate": 8.031208167400833e-06, | |
| "loss": 0.0194, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 4.222972972972973, | |
| "grad_norm": 0.4366669926231541, | |
| "learning_rate": 7.906415589848834e-06, | |
| "loss": 0.0175, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.239864864864865, | |
| "grad_norm": 0.4875801354190107, | |
| "learning_rate": 7.78406857322756e-06, | |
| "loss": 0.0176, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 4.256756756756757, | |
| "grad_norm": 0.5496320514624984, | |
| "learning_rate": 7.664182388203037e-06, | |
| "loss": 0.0183, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.273648648648648, | |
| "grad_norm": 0.4131315904135289, | |
| "learning_rate": 7.54677199829414e-06, | |
| "loss": 0.0175, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 4.29054054054054, | |
| "grad_norm": 0.5104334470158947, | |
| "learning_rate": 7.4318520580049444e-06, | |
| "loss": 0.0173, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.3074324324324325, | |
| "grad_norm": 0.5028181787292044, | |
| "learning_rate": 7.3194369109956e-06, | |
| "loss": 0.0193, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.324324324324325, | |
| "grad_norm": 0.42628529148210376, | |
| "learning_rate": 7.209540588292083e-06, | |
| "loss": 0.0158, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.341216216216216, | |
| "grad_norm": 0.44913562810367375, | |
| "learning_rate": 7.102176806534873e-06, | |
| "loss": 0.0177, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 4.358108108108108, | |
| "grad_norm": 0.5514097182218852, | |
| "learning_rate": 6.9973589662669455e-06, | |
| "loss": 0.017, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.375, | |
| "grad_norm": 0.5163780905038693, | |
| "learning_rate": 6.8951001502612065e-06, | |
| "loss": 0.0168, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 4.391891891891892, | |
| "grad_norm": 0.49290296808784145, | |
| "learning_rate": 6.7954131218875404e-06, | |
| "loss": 0.016, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.408783783783784, | |
| "grad_norm": 0.430553207351477, | |
| "learning_rate": 6.69831032351977e-06, | |
| "loss": 0.0169, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 4.425675675675675, | |
| "grad_norm": 0.4642297721283167, | |
| "learning_rate": 6.603803874982687e-06, | |
| "loss": 0.0173, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.4425675675675675, | |
| "grad_norm": 0.41074319788773356, | |
| "learning_rate": 6.511905572039298e-06, | |
| "loss": 0.0191, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 4.45945945945946, | |
| "grad_norm": 0.42290407551524545, | |
| "learning_rate": 6.422626884918559e-06, | |
| "loss": 0.0187, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.476351351351352, | |
| "grad_norm": 0.47394273027675116, | |
| "learning_rate": 6.33597895688373e-06, | |
| "loss": 0.0173, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.493243243243243, | |
| "grad_norm": 0.4957247279877308, | |
| "learning_rate": 6.2519726028415145e-06, | |
| "loss": 0.0176, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.510135135135135, | |
| "grad_norm": 0.40363156811514256, | |
| "learning_rate": 6.170618307992231e-06, | |
| "loss": 0.0156, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 4.527027027027027, | |
| "grad_norm": 0.46394127923496636, | |
| "learning_rate": 6.091926226521089e-06, | |
| "loss": 0.0166, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.543918918918919, | |
| "grad_norm": 0.4892778778622955, | |
| "learning_rate": 6.015906180330808e-06, | |
| "loss": 0.016, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 4.5608108108108105, | |
| "grad_norm": 0.39032273135281653, | |
| "learning_rate": 5.942567657815696e-06, | |
| "loss": 0.0168, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.577702702702703, | |
| "grad_norm": 0.4719537909270886, | |
| "learning_rate": 5.871919812677383e-06, | |
| "loss": 0.0164, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 4.594594594594595, | |
| "grad_norm": 0.4245185329209189, | |
| "learning_rate": 5.8039714627822754e-06, | |
| "loss": 0.0156, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.611486486486487, | |
| "grad_norm": 0.41475267241989133, | |
| "learning_rate": 5.738731089060995e-06, | |
| "loss": 0.0153, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 4.628378378378378, | |
| "grad_norm": 0.5352228892671703, | |
| "learning_rate": 5.676206834449797e-06, | |
| "loss": 0.0165, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.64527027027027, | |
| "grad_norm": 0.37701331899842194, | |
| "learning_rate": 5.616406502874251e-06, | |
| "loss": 0.0146, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.662162162162162, | |
| "grad_norm": 0.39899803271574863, | |
| "learning_rate": 5.55933755827518e-06, | |
| "loss": 0.0148, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.679054054054054, | |
| "grad_norm": 0.48339475577967467, | |
| "learning_rate": 5.505007123677063e-06, | |
| "loss": 0.0161, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 4.695945945945946, | |
| "grad_norm": 0.47920447606296807, | |
| "learning_rate": 5.453421980298957e-06, | |
| "loss": 0.0161, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.712837837837838, | |
| "grad_norm": 0.45349300791214964, | |
| "learning_rate": 5.4045885667081375e-06, | |
| "loss": 0.0162, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 4.72972972972973, | |
| "grad_norm": 0.4524858740954348, | |
| "learning_rate": 5.358512978016445e-06, | |
| "loss": 0.0156, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.746621621621622, | |
| "grad_norm": 0.4052610898199223, | |
| "learning_rate": 5.315200965119541e-06, | |
| "loss": 0.0144, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 4.763513513513513, | |
| "grad_norm": 0.4648951449721674, | |
| "learning_rate": 5.27465793397911e-06, | |
| "loss": 0.0174, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.780405405405405, | |
| "grad_norm": 0.4310666293350422, | |
| "learning_rate": 5.236888944948117e-06, | |
| "loss": 0.0157, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 4.797297297297297, | |
| "grad_norm": 0.4423747920595358, | |
| "learning_rate": 5.201898712139201e-06, | |
| "loss": 0.0139, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.8141891891891895, | |
| "grad_norm": 0.40535469669136887, | |
| "learning_rate": 5.1696916028362964e-06, | |
| "loss": 0.0161, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 4.831081081081081, | |
| "grad_norm": 0.5159753334356647, | |
| "learning_rate": 5.1402716369495194e-06, | |
| "loss": 0.0172, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.847972972972973, | |
| "grad_norm": 0.49504962870437985, | |
| "learning_rate": 5.113642486513428e-06, | |
| "loss": 0.0148, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 4.864864864864865, | |
| "grad_norm": 0.4125316130138178, | |
| "learning_rate": 5.089807475228711e-06, | |
| "loss": 0.0142, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.881756756756757, | |
| "grad_norm": 0.4694139851875935, | |
| "learning_rate": 5.06876957804733e-06, | |
| "loss": 0.0157, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 4.898648648648649, | |
| "grad_norm": 0.44686436058551265, | |
| "learning_rate": 5.050531420801205e-06, | |
| "loss": 0.0159, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.91554054054054, | |
| "grad_norm": 0.4838100458198807, | |
| "learning_rate": 5.03509527987448e-06, | |
| "loss": 0.0164, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 4.9324324324324325, | |
| "grad_norm": 0.44838862569063753, | |
| "learning_rate": 5.022463081919386e-06, | |
| "loss": 0.0135, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.949324324324325, | |
| "grad_norm": 0.33222610870493824, | |
| "learning_rate": 5.012636403615775e-06, | |
| "loss": 0.0157, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 4.966216216216216, | |
| "grad_norm": 0.4899278481273953, | |
| "learning_rate": 5.005616471474332e-06, | |
| "loss": 0.015, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.983108108108108, | |
| "grad_norm": 0.46395409247351443, | |
| "learning_rate": 5.001404161683473e-06, | |
| "loss": 0.0157, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.38146340884683544, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0141, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1480, | |
| "total_flos": 605863668678656.0, | |
| "train_loss": 0.36788542065266017, | |
| "train_runtime": 21782.9524, | |
| "train_samples_per_second": 2.171, | |
| "train_steps_per_second": 0.068 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 605863668678656.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |