| { | |
| "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": 6.462106284368249, | |
| "learning_rate": 3.3783783783783788e-06, | |
| "loss": 1.2799, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.033783783783783786, | |
| "grad_norm": 5.796465322090743, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 1.1537, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05067567567567568, | |
| "grad_norm": 4.2218820593408095, | |
| "learning_rate": 1.0135135135135136e-05, | |
| "loss": 1.0065, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06756756756756757, | |
| "grad_norm": 3.7249576431741684, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 0.9447, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08445945945945946, | |
| "grad_norm": 3.190895360133061, | |
| "learning_rate": 1.6891891891891892e-05, | |
| "loss": 0.9273, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.10135135135135136, | |
| "grad_norm": 2.970371594299886, | |
| "learning_rate": 2.0270270270270273e-05, | |
| "loss": 0.9025, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11824324324324324, | |
| "grad_norm": 2.963735798525872, | |
| "learning_rate": 2.364864864864865e-05, | |
| "loss": 0.9132, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.13513513513513514, | |
| "grad_norm": 2.8827468495825426, | |
| "learning_rate": 2.702702702702703e-05, | |
| "loss": 0.9218, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15202702702702703, | |
| "grad_norm": 2.761334881573851, | |
| "learning_rate": 3.0405405405405407e-05, | |
| "loss": 0.9476, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16891891891891891, | |
| "grad_norm": 2.668519821436992, | |
| "learning_rate": 3.3783783783783784e-05, | |
| "loss": 0.9359, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1858108108108108, | |
| "grad_norm": 2.6190244017441064, | |
| "learning_rate": 3.7162162162162165e-05, | |
| "loss": 0.9183, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.20270270270270271, | |
| "grad_norm": 2.719864590265737, | |
| "learning_rate": 4.0540540540540545e-05, | |
| "loss": 0.9458, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2195945945945946, | |
| "grad_norm": 2.4818549466218336, | |
| "learning_rate": 4.391891891891892e-05, | |
| "loss": 0.9429, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.23648648648648649, | |
| "grad_norm": 2.724969057424987, | |
| "learning_rate": 4.72972972972973e-05, | |
| "loss": 0.9683, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2533783783783784, | |
| "grad_norm": 2.326953092679936, | |
| "learning_rate": 4.999994383297182e-05, | |
| "loss": 0.9661, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 2.4334462522009828, | |
| "learning_rate": 4.9997978016429605e-05, | |
| "loss": 0.9672, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.28716216216216217, | |
| "grad_norm": 2.566697711672114, | |
| "learning_rate": 4.9993204128893056e-05, | |
| "loss": 0.9751, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.30405405405405406, | |
| "grad_norm": 2.044279421214079, | |
| "learning_rate": 4.9985622766211935e-05, | |
| "loss": 0.9376, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.32094594594594594, | |
| "grad_norm": 2.094911920965702, | |
| "learning_rate": 4.997523487464928e-05, | |
| "loss": 0.9416, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.33783783783783783, | |
| "grad_norm": 2.150093740481432, | |
| "learning_rate": 4.996204175076325e-05, | |
| "loss": 0.9629, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3547297297297297, | |
| "grad_norm": 2.136510625512231, | |
| "learning_rate": 4.99460450412453e-05, | |
| "loss": 0.9651, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3716216216216216, | |
| "grad_norm": 1.9860540005802636, | |
| "learning_rate": 4.99272467427147e-05, | |
| "loss": 0.9677, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3885135135135135, | |
| "grad_norm": 2.1097510332573397, | |
| "learning_rate": 4.990564920146926e-05, | |
| "loss": 0.9881, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.40540540540540543, | |
| "grad_norm": 2.0395479749604557, | |
| "learning_rate": 4.9881255113192526e-05, | |
| "loss": 0.9276, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4222972972972973, | |
| "grad_norm": 2.13291160788017, | |
| "learning_rate": 4.985406752261731e-05, | |
| "loss": 0.9456, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4391891891891892, | |
| "grad_norm": 1.9035262700170335, | |
| "learning_rate": 4.982408982314565e-05, | |
| "loss": 0.9681, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4560810810810811, | |
| "grad_norm": 2.169802087902646, | |
| "learning_rate": 4.9791325756425264e-05, | |
| "loss": 0.9418, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.47297297297297297, | |
| "grad_norm": 1.693828140694662, | |
| "learning_rate": 4.975577941188258e-05, | |
| "loss": 0.9342, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.48986486486486486, | |
| "grad_norm": 1.8426685577613464, | |
| "learning_rate": 4.971745522621225e-05, | |
| "loss": 0.9583, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.5067567567567568, | |
| "grad_norm": 1.7092078037372602, | |
| "learning_rate": 4.967635798282344e-05, | |
| "loss": 0.975, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5236486486486487, | |
| "grad_norm": 1.8143108389578015, | |
| "learning_rate": 4.963249281124278e-05, | |
| "loss": 0.9144, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 1.6486030826900318, | |
| "learning_rate": 4.958586518647407e-05, | |
| "loss": 0.904, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5574324324324325, | |
| "grad_norm": 1.581894032306616, | |
| "learning_rate": 4.953648092831505e-05, | |
| "loss": 0.9359, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5743243243243243, | |
| "grad_norm": 1.7011324331322406, | |
| "learning_rate": 4.9484346200630855e-05, | |
| "loss": 0.9401, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5912162162162162, | |
| "grad_norm": 1.7197831463163102, | |
| "learning_rate": 4.9429467510584794e-05, | |
| "loss": 0.8993, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6081081081081081, | |
| "grad_norm": 1.7969747554181748, | |
| "learning_rate": 4.937185170782607e-05, | |
| "loss": 0.941, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 1.7326030021356258, | |
| "learning_rate": 4.931150598363494e-05, | |
| "loss": 0.9136, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6418918918918919, | |
| "grad_norm": 1.6989345371043831, | |
| "learning_rate": 4.9248437870025035e-05, | |
| "loss": 0.9027, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6587837837837838, | |
| "grad_norm": 1.7071249886299835, | |
| "learning_rate": 4.9182655238803365e-05, | |
| "loss": 0.9028, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 1.6364386352714728, | |
| "learning_rate": 4.911416630058772e-05, | |
| "loss": 0.8891, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6925675675675675, | |
| "grad_norm": 1.6826896509560934, | |
| "learning_rate": 4.9042979603781884e-05, | |
| "loss": 0.9066, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.7094594594594594, | |
| "grad_norm": 1.5845413735256957, | |
| "learning_rate": 4.896910403350873e-05, | |
| "loss": 0.9229, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7263513513513513, | |
| "grad_norm": 1.651057950784663, | |
| "learning_rate": 4.8892548810501146e-05, | |
| "loss": 0.9044, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7432432432432432, | |
| "grad_norm": 1.504806433052808, | |
| "learning_rate": 4.88133234899512e-05, | |
| "loss": 0.8887, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7601351351351351, | |
| "grad_norm": 1.6483855326386294, | |
| "learning_rate": 4.873143796031752e-05, | |
| "loss": 0.8843, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.777027027027027, | |
| "grad_norm": 1.5281476898536954, | |
| "learning_rate": 4.864690244209105e-05, | |
| "loss": 0.8905, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.793918918918919, | |
| "grad_norm": 1.6270015828836966, | |
| "learning_rate": 4.855972748651939e-05, | |
| "loss": 0.9052, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 1.570164320094409, | |
| "learning_rate": 4.8469923974289874e-05, | |
| "loss": 0.8783, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8277027027027027, | |
| "grad_norm": 1.7197408315182336, | |
| "learning_rate": 4.837750311417146e-05, | |
| "loss": 0.8776, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8445945945945946, | |
| "grad_norm": 1.706044145062568, | |
| "learning_rate": 4.828247644161577e-05, | |
| "loss": 0.9271, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8614864864864865, | |
| "grad_norm": 1.615230215894314, | |
| "learning_rate": 4.8184855817317226e-05, | |
| "loss": 0.8851, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8783783783783784, | |
| "grad_norm": 1.3941648443551302, | |
| "learning_rate": 4.808465342573274e-05, | |
| "loss": 0.8629, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8952702702702703, | |
| "grad_norm": 1.7444726060654847, | |
| "learning_rate": 4.7981881773560886e-05, | |
| "loss": 0.8933, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.9121621621621622, | |
| "grad_norm": 1.4855383267727116, | |
| "learning_rate": 4.787655368818087e-05, | |
| "loss": 0.8642, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9290540540540541, | |
| "grad_norm": 1.5235974901150469, | |
| "learning_rate": 4.77686823160515e-05, | |
| "loss": 0.8646, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9459459459459459, | |
| "grad_norm": 1.5469261416890907, | |
| "learning_rate": 4.765828112107034e-05, | |
| "loss": 0.8691, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9628378378378378, | |
| "grad_norm": 1.477007681771187, | |
| "learning_rate": 4.75453638828932e-05, | |
| "loss": 0.8656, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9797297297297297, | |
| "grad_norm": 1.4326160380137902, | |
| "learning_rate": 4.742994469521421e-05, | |
| "loss": 0.8594, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9966216216216216, | |
| "grad_norm": 1.3672369112488894, | |
| "learning_rate": 4.7312037964006806e-05, | |
| "loss": 0.8542, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.0135135135135136, | |
| "grad_norm": 1.7102173891202228, | |
| "learning_rate": 4.719165840572557e-05, | |
| "loss": 0.6417, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0304054054054055, | |
| "grad_norm": 1.352890869957888, | |
| "learning_rate": 4.7068821045469464e-05, | |
| "loss": 0.5555, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0472972972972974, | |
| "grad_norm": 1.6709854946304417, | |
| "learning_rate": 4.694354121510644e-05, | |
| "loss": 0.5489, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0641891891891893, | |
| "grad_norm": 1.5112125205457065, | |
| "learning_rate": 4.681583455135985e-05, | |
| "loss": 0.5353, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0810810810810811, | |
| "grad_norm": 1.3624666862976014, | |
| "learning_rate": 4.668571699385668e-05, | |
| "loss": 0.5432, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.097972972972973, | |
| "grad_norm": 1.5384986763994564, | |
| "learning_rate": 4.655320478313817e-05, | |
| "loss": 0.5344, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.114864864864865, | |
| "grad_norm": 1.3995917440377128, | |
| "learning_rate": 4.641831445863265e-05, | |
| "loss": 0.5531, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.1317567567567568, | |
| "grad_norm": 1.4884708367489115, | |
| "learning_rate": 4.628106285659124e-05, | |
| "loss": 0.5705, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1486486486486487, | |
| "grad_norm": 1.766828873521366, | |
| "learning_rate": 4.614146710798645e-05, | |
| "loss": 0.5437, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1655405405405406, | |
| "grad_norm": 1.4872551854409901, | |
| "learning_rate": 4.599954463637394e-05, | |
| "loss": 0.5852, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1824324324324325, | |
| "grad_norm": 1.3549745842300658, | |
| "learning_rate": 4.585531315571788e-05, | |
| "loss": 0.5446, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1993243243243243, | |
| "grad_norm": 1.4438381968118856, | |
| "learning_rate": 4.570879066817991e-05, | |
| "loss": 0.5498, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.2162162162162162, | |
| "grad_norm": 1.4828526458644822, | |
| "learning_rate": 4.555999546187229e-05, | |
| "loss": 0.5718, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.2331081081081081, | |
| "grad_norm": 1.4194022003950522, | |
| "learning_rate": 4.5408946108575215e-05, | |
| "loss": 0.5636, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.3621643995050003, | |
| "learning_rate": 4.5255661461418854e-05, | |
| "loss": 0.5568, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.2668918918918919, | |
| "grad_norm": 1.463520965102366, | |
| "learning_rate": 4.510016065253016e-05, | |
| "loss": 0.5558, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2837837837837838, | |
| "grad_norm": 1.4624357397789567, | |
| "learning_rate": 4.4942463090644896e-05, | |
| "loss": 0.5892, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.3006756756756757, | |
| "grad_norm": 1.3917330520100684, | |
| "learning_rate": 4.478258845868522e-05, | |
| "loss": 0.5735, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.3175675675675675, | |
| "grad_norm": 1.4054292296666042, | |
| "learning_rate": 4.462055671130289e-05, | |
| "loss": 0.5699, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.3344594594594594, | |
| "grad_norm": 1.5019153889843935, | |
| "learning_rate": 4.445638807238865e-05, | |
| "loss": 0.5738, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 1.5369414617688213, | |
| "learning_rate": 4.4290103032548094e-05, | |
| "loss": 0.5679, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.3682432432432432, | |
| "grad_norm": 1.2808436044425764, | |
| "learning_rate": 4.412172234654399e-05, | |
| "loss": 0.5483, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.385135135135135, | |
| "grad_norm": 1.481756459621886, | |
| "learning_rate": 4.395126703070589e-05, | |
| "loss": 0.5503, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.402027027027027, | |
| "grad_norm": 1.430885737073211, | |
| "learning_rate": 4.3778758360306986e-05, | |
| "loss": 0.5638, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.4189189189189189, | |
| "grad_norm": 1.2997035063938607, | |
| "learning_rate": 4.360421786690862e-05, | |
| "loss": 0.5535, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.4358108108108107, | |
| "grad_norm": 1.4154480312272155, | |
| "learning_rate": 4.3427667335672815e-05, | |
| "loss": 0.5732, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4527027027027026, | |
| "grad_norm": 1.3832906340350364, | |
| "learning_rate": 4.324912880264326e-05, | |
| "loss": 0.5452, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.4695945945945945, | |
| "grad_norm": 1.4379805126978302, | |
| "learning_rate": 4.30686245519948e-05, | |
| "loss": 0.5404, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.4864864864864864, | |
| "grad_norm": 1.2425338062053777, | |
| "learning_rate": 4.288617711325207e-05, | |
| "loss": 0.5816, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.5033783783783785, | |
| "grad_norm": 1.4500671971300003, | |
| "learning_rate": 4.2701809258477575e-05, | |
| "loss": 0.5762, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.5202702702702702, | |
| "grad_norm": 1.2990880876677995, | |
| "learning_rate": 4.251554399942928e-05, | |
| "loss": 0.565, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.5371621621621623, | |
| "grad_norm": 1.2337100226741404, | |
| "learning_rate": 4.2327404584688495e-05, | |
| "loss": 0.589, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.554054054054054, | |
| "grad_norm": 1.4390554313953952, | |
| "learning_rate": 4.21374144967581e-05, | |
| "loss": 0.5575, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.570945945945946, | |
| "grad_norm": 1.430055972450592, | |
| "learning_rate": 4.194559744913157e-05, | |
| "loss": 0.556, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.5878378378378377, | |
| "grad_norm": 1.380773353267627, | |
| "learning_rate": 4.1751977383333224e-05, | |
| "loss": 0.5558, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.6047297297297298, | |
| "grad_norm": 1.3482546972942344, | |
| "learning_rate": 4.1556578465929955e-05, | |
| "loss": 0.5638, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.6216216216216215, | |
| "grad_norm": 1.3888644742506622, | |
| "learning_rate": 4.1359425085514906e-05, | |
| "loss": 0.5857, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.6385135135135136, | |
| "grad_norm": 1.3257552271684938, | |
| "learning_rate": 4.1160541849663404e-05, | |
| "loss": 0.5686, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.6554054054054053, | |
| "grad_norm": 1.3125538494518993, | |
| "learning_rate": 4.095995358186162e-05, | |
| "loss": 0.5698, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.6722972972972974, | |
| "grad_norm": 1.2858747479904953, | |
| "learning_rate": 4.0757685318408186e-05, | |
| "loss": 0.5505, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.689189189189189, | |
| "grad_norm": 1.2802134094898805, | |
| "learning_rate": 4.055376230528936e-05, | |
| "loss": 0.5576, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.7060810810810811, | |
| "grad_norm": 1.354705799711518, | |
| "learning_rate": 4.034820999502794e-05, | |
| "loss": 0.5684, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.722972972972973, | |
| "grad_norm": 1.2611870103615626, | |
| "learning_rate": 4.0141054043506406e-05, | |
| "loss": 0.5705, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.739864864864865, | |
| "grad_norm": 1.2589476932660713, | |
| "learning_rate": 3.993232030676473e-05, | |
| "loss": 0.5635, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.7567567567567568, | |
| "grad_norm": 1.2999787013352957, | |
| "learning_rate": 3.972203483777315e-05, | |
| "loss": 0.5509, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.7736486486486487, | |
| "grad_norm": 1.4037046064106253, | |
| "learning_rate": 3.95102238831804e-05, | |
| "loss": 0.5898, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.7905405405405406, | |
| "grad_norm": 1.3826344219389248, | |
| "learning_rate": 3.929691388003772e-05, | |
| "loss": 0.567, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.8074324324324325, | |
| "grad_norm": 1.325240320783927, | |
| "learning_rate": 3.908213145249916e-05, | |
| "loss": 0.5653, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.8243243243243243, | |
| "grad_norm": 1.329132097993209, | |
| "learning_rate": 3.886590340849852e-05, | |
| "loss": 0.564, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.8412162162162162, | |
| "grad_norm": 1.3619617065732015, | |
| "learning_rate": 3.864825673640326e-05, | |
| "loss": 0.5512, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.8581081081081081, | |
| "grad_norm": 1.3104136200516292, | |
| "learning_rate": 3.842921860164607e-05, | |
| "loss": 0.562, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 1.3568591849701919, | |
| "learning_rate": 3.8208816343334156e-05, | |
| "loss": 0.5474, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.8918918918918919, | |
| "grad_norm": 1.3029059698575418, | |
| "learning_rate": 3.798707747083694e-05, | |
| "loss": 0.57, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.9087837837837838, | |
| "grad_norm": 1.3161190691167883, | |
| "learning_rate": 3.776402966035251e-05, | |
| "loss": 0.5763, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.9256756756756757, | |
| "grad_norm": 1.4737785688634046, | |
| "learning_rate": 3.753970075145322e-05, | |
| "loss": 0.5482, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.9425675675675675, | |
| "grad_norm": 1.3302662116752713, | |
| "learning_rate": 3.731411874361094e-05, | |
| "loss": 0.5425, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.9594594594594594, | |
| "grad_norm": 1.2988092591143818, | |
| "learning_rate": 3.7087311792702265e-05, | |
| "loss": 0.5587, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.9763513513513513, | |
| "grad_norm": 1.3706395491828778, | |
| "learning_rate": 3.685930820749433e-05, | |
| "loss": 0.5518, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.9932432432432432, | |
| "grad_norm": 1.2498855644934173, | |
| "learning_rate": 3.663013644611139e-05, | |
| "loss": 0.5511, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.010135135135135, | |
| "grad_norm": 1.182047436740319, | |
| "learning_rate": 3.639982511248289e-05, | |
| "loss": 0.3749, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.027027027027027, | |
| "grad_norm": 1.292806145449219, | |
| "learning_rate": 3.616840295277328e-05, | |
| "loss": 0.2401, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.043918918918919, | |
| "grad_norm": 1.2836873422540378, | |
| "learning_rate": 3.593589885179405e-05, | |
| "loss": 0.2507, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.060810810810811, | |
| "grad_norm": 1.1629236979557103, | |
| "learning_rate": 3.5702341829398525e-05, | |
| "loss": 0.2436, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.0777027027027026, | |
| "grad_norm": 1.2145870883157541, | |
| "learning_rate": 3.5467761036859736e-05, | |
| "loss": 0.2473, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.0945945945945947, | |
| "grad_norm": 1.1669700401553293, | |
| "learning_rate": 3.523218575323198e-05, | |
| "loss": 0.2395, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.1114864864864864, | |
| "grad_norm": 1.0949827165418597, | |
| "learning_rate": 3.499564538169629e-05, | |
| "loss": 0.2386, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.1283783783783785, | |
| "grad_norm": 1.2354405531271948, | |
| "learning_rate": 3.475816944589058e-05, | |
| "loss": 0.2438, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.14527027027027, | |
| "grad_norm": 1.2220320680113108, | |
| "learning_rate": 3.451978758622458e-05, | |
| "loss": 0.2554, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.1621621621621623, | |
| "grad_norm": 1.1171635389193193, | |
| "learning_rate": 3.4280529556180404e-05, | |
| "loss": 0.2426, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.179054054054054, | |
| "grad_norm": 1.1397501569198598, | |
| "learning_rate": 3.4040425218598755e-05, | |
| "loss": 0.2406, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.195945945945946, | |
| "grad_norm": 1.2436627285985906, | |
| "learning_rate": 3.379950454195172e-05, | |
| "loss": 0.2498, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.2128378378378377, | |
| "grad_norm": 1.177199277178587, | |
| "learning_rate": 3.355779759660223e-05, | |
| "loss": 0.2468, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.22972972972973, | |
| "grad_norm": 1.2056688245242106, | |
| "learning_rate": 3.331533455105084e-05, | |
| "loss": 0.2398, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.2466216216216215, | |
| "grad_norm": 1.1195023322495838, | |
| "learning_rate": 3.307214566817027e-05, | |
| "loss": 0.2477, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.2635135135135136, | |
| "grad_norm": 1.158884403290133, | |
| "learning_rate": 3.2828261301428206e-05, | |
| "loss": 0.251, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.2804054054054053, | |
| "grad_norm": 1.163288318918253, | |
| "learning_rate": 3.2583711891098665e-05, | |
| "loss": 0.2521, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.2972972972972974, | |
| "grad_norm": 1.2145258921719575, | |
| "learning_rate": 3.23385279604627e-05, | |
| "loss": 0.2537, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.314189189189189, | |
| "grad_norm": 1.180178858453737, | |
| "learning_rate": 3.209274011199861e-05, | |
| "loss": 0.251, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.331081081081081, | |
| "grad_norm": 1.1237237084221496, | |
| "learning_rate": 3.18463790235623e-05, | |
| "loss": 0.2526, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.347972972972973, | |
| "grad_norm": 1.2239335542726613, | |
| "learning_rate": 3.159947544455828e-05, | |
| "loss": 0.2469, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.364864864864865, | |
| "grad_norm": 1.2230291299298848, | |
| "learning_rate": 3.135206019210167e-05, | |
| "loss": 0.2481, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3817567567567566, | |
| "grad_norm": 1.113596170808197, | |
| "learning_rate": 3.110416414717181e-05, | |
| "loss": 0.2583, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.3986486486486487, | |
| "grad_norm": 1.1731188839672357, | |
| "learning_rate": 3.085581825075782e-05, | |
| "loss": 0.2431, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.4155405405405403, | |
| "grad_norm": 1.1790462900350855, | |
| "learning_rate": 3.060705349999677e-05, | |
| "loss": 0.2593, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.4324324324324325, | |
| "grad_norm": 1.1562129145645026, | |
| "learning_rate": 3.0357900944304774e-05, | |
| "loss": 0.2552, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.4493243243243246, | |
| "grad_norm": 1.079877803058021, | |
| "learning_rate": 3.0108391681501564e-05, | |
| "loss": 0.2479, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.4662162162162162, | |
| "grad_norm": 1.1505133985264928, | |
| "learning_rate": 2.9858556853929048e-05, | |
| "loss": 0.2546, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.483108108108108, | |
| "grad_norm": 1.112481855217279, | |
| "learning_rate": 2.96084276445643e-05, | |
| "loss": 0.2598, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.1013323666381978, | |
| "learning_rate": 2.9358035273127483e-05, | |
| "loss": 0.241, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.516891891891892, | |
| "grad_norm": 1.15739883490395, | |
| "learning_rate": 2.910741099218514e-05, | |
| "loss": 0.251, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 2.5337837837837838, | |
| "grad_norm": 1.2252184761757072, | |
| "learning_rate": 2.8856586083249487e-05, | |
| "loss": 0.2574, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.5506756756756754, | |
| "grad_norm": 1.20081050339093, | |
| "learning_rate": 2.860559185287397e-05, | |
| "loss": 0.2492, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 2.5675675675675675, | |
| "grad_norm": 1.1421908058638464, | |
| "learning_rate": 2.83544596287458e-05, | |
| "loss": 0.25, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.5844594594594597, | |
| "grad_norm": 1.2024371718558493, | |
| "learning_rate": 2.8103220755775776e-05, | |
| "loss": 0.2507, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 2.6013513513513513, | |
| "grad_norm": 1.1267137517129018, | |
| "learning_rate": 2.785190659218604e-05, | |
| "loss": 0.2556, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.618243243243243, | |
| "grad_norm": 1.236903788139508, | |
| "learning_rate": 2.760054850559603e-05, | |
| "loss": 0.2433, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.635135135135135, | |
| "grad_norm": 1.2356894950363584, | |
| "learning_rate": 2.7349177869107462e-05, | |
| "loss": 0.2438, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.652027027027027, | |
| "grad_norm": 1.1854186446919905, | |
| "learning_rate": 2.709782605738842e-05, | |
| "loss": 0.2505, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 2.668918918918919, | |
| "grad_norm": 1.1863404508336286, | |
| "learning_rate": 2.684652444275741e-05, | |
| "loss": 0.2481, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.685810810810811, | |
| "grad_norm": 1.287742827707764, | |
| "learning_rate": 2.6595304391267605e-05, | |
| "loss": 0.246, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 1.1464548020357286, | |
| "learning_rate": 2.634419725879193e-05, | |
| "loss": 0.24, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.7195945945945947, | |
| "grad_norm": 1.1726381757168638, | |
| "learning_rate": 2.60932343871094e-05, | |
| "loss": 0.2369, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 2.7364864864864864, | |
| "grad_norm": 1.2183808849866478, | |
| "learning_rate": 2.58424470999932e-05, | |
| "loss": 0.2399, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.7533783783783785, | |
| "grad_norm": 1.1780764773888983, | |
| "learning_rate": 2.5591866699301055e-05, | |
| "loss": 0.238, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 2.77027027027027, | |
| "grad_norm": 1.1405655258294773, | |
| "learning_rate": 2.534152446106825e-05, | |
| "loss": 0.2417, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.7871621621621623, | |
| "grad_norm": 1.2008872094756395, | |
| "learning_rate": 2.5091451631604033e-05, | |
| "loss": 0.2403, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.804054054054054, | |
| "grad_norm": 1.2336776534307317, | |
| "learning_rate": 2.4841679423591523e-05, | |
| "loss": 0.2385, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.820945945945946, | |
| "grad_norm": 1.1106811690944616, | |
| "learning_rate": 2.4592239012191977e-05, | |
| "loss": 0.2348, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.8378378378378377, | |
| "grad_norm": 1.0383736357246989, | |
| "learning_rate": 2.4343161531153647e-05, | |
| "loss": 0.2318, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.85472972972973, | |
| "grad_norm": 1.170166493718478, | |
| "learning_rate": 2.4094478068925884e-05, | |
| "loss": 0.2489, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.8716216216216215, | |
| "grad_norm": 1.1394443830669667, | |
| "learning_rate": 2.3846219664778824e-05, | |
| "loss": 0.2327, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.8885135135135136, | |
| "grad_norm": 1.0682485518263032, | |
| "learning_rate": 2.3598417304929226e-05, | |
| "loss": 0.2366, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.9054054054054053, | |
| "grad_norm": 1.1539830932451005, | |
| "learning_rate": 2.3351101918672985e-05, | |
| "loss": 0.2317, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.9222972972972974, | |
| "grad_norm": 1.1643714559592864, | |
| "learning_rate": 2.3104304374524704e-05, | |
| "loss": 0.2345, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.939189189189189, | |
| "grad_norm": 1.1872533494513688, | |
| "learning_rate": 2.2858055476364822e-05, | |
| "loss": 0.2428, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.956081081081081, | |
| "grad_norm": 1.1078526680690635, | |
| "learning_rate": 2.2612385959594877e-05, | |
| "loss": 0.2325, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.972972972972973, | |
| "grad_norm": 1.1173105295894745, | |
| "learning_rate": 2.2367326487301317e-05, | |
| "loss": 0.2377, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.989864864864865, | |
| "grad_norm": 1.146990785070699, | |
| "learning_rate": 2.2122907646428214e-05, | |
| "loss": 0.2322, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.0067567567567566, | |
| "grad_norm": 0.7760183399891705, | |
| "learning_rate": 2.1879159943959686e-05, | |
| "loss": 0.1711, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.0236486486486487, | |
| "grad_norm": 0.7976649358102162, | |
| "learning_rate": 2.1636113803112097e-05, | |
| "loss": 0.0884, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.0405405405405403, | |
| "grad_norm": 1.0091290097140033, | |
| "learning_rate": 2.139379955953686e-05, | |
| "loss": 0.0836, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.0574324324324325, | |
| "grad_norm": 0.9660423218316871, | |
| "learning_rate": 2.1152247457534065e-05, | |
| "loss": 0.0823, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.074324324324324, | |
| "grad_norm": 0.9235201506843036, | |
| "learning_rate": 2.0911487646277623e-05, | |
| "loss": 0.0834, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.0912162162162162, | |
| "grad_norm": 0.824557453248684, | |
| "learning_rate": 2.067155017605212e-05, | |
| "loss": 0.082, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.108108108108108, | |
| "grad_norm": 0.8414113561244017, | |
| "learning_rate": 2.0432464994502203e-05, | |
| "loss": 0.0836, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 0.8910092987005708, | |
| "learning_rate": 2.0194261942894628e-05, | |
| "loss": 0.0856, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.141891891891892, | |
| "grad_norm": 0.8672585195826027, | |
| "learning_rate": 1.995697075239365e-05, | |
| "loss": 0.0842, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.1587837837837838, | |
| "grad_norm": 0.845424189240467, | |
| "learning_rate": 1.972062104035017e-05, | |
| "loss": 0.0823, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.175675675675676, | |
| "grad_norm": 0.8566673924138892, | |
| "learning_rate": 1.9485242306605028e-05, | |
| "loss": 0.0908, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.1925675675675675, | |
| "grad_norm": 0.8200502766583625, | |
| "learning_rate": 1.9250863929807027e-05, | |
| "loss": 0.0777, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.2094594594594597, | |
| "grad_norm": 0.9603217468544512, | |
| "learning_rate": 1.9017515163746058e-05, | |
| "loss": 0.0824, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.2263513513513513, | |
| "grad_norm": 0.7892710086022029, | |
| "learning_rate": 1.878522513370177e-05, | |
| "loss": 0.083, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.2432432432432434, | |
| "grad_norm": 0.8686350772681539, | |
| "learning_rate": 1.855402283280836e-05, | |
| "loss": 0.081, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.260135135135135, | |
| "grad_norm": 0.8007518648810419, | |
| "learning_rate": 1.8323937118435786e-05, | |
| "loss": 0.0821, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.277027027027027, | |
| "grad_norm": 0.9003546198296242, | |
| "learning_rate": 1.8094996708587958e-05, | |
| "loss": 0.0863, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.293918918918919, | |
| "grad_norm": 1.0054385141581792, | |
| "learning_rate": 1.7867230178318334e-05, | |
| "loss": 0.0794, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.310810810810811, | |
| "grad_norm": 0.8872527151741653, | |
| "learning_rate": 1.7640665956163306e-05, | |
| "loss": 0.0797, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.3277027027027026, | |
| "grad_norm": 0.7977119970077569, | |
| "learning_rate": 1.7415332320593964e-05, | |
| "loss": 0.079, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.3445945945945947, | |
| "grad_norm": 0.8448885628803706, | |
| "learning_rate": 1.719125739648648e-05, | |
| "loss": 0.0808, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 3.3614864864864864, | |
| "grad_norm": 0.8453887264829997, | |
| "learning_rate": 1.6968469151611766e-05, | |
| "loss": 0.077, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "grad_norm": 0.8136095971332963, | |
| "learning_rate": 1.6746995393144668e-05, | |
| "loss": 0.079, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.39527027027027, | |
| "grad_norm": 0.9218687872899989, | |
| "learning_rate": 1.6526863764193228e-05, | |
| "loss": 0.08, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 3.4121621621621623, | |
| "grad_norm": 0.8502918546022624, | |
| "learning_rate": 1.6308101740348433e-05, | |
| "loss": 0.0849, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 3.429054054054054, | |
| "grad_norm": 0.7936393521711613, | |
| "learning_rate": 1.6090736626254894e-05, | |
| "loss": 0.08, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 3.445945945945946, | |
| "grad_norm": 0.8187403520204729, | |
| "learning_rate": 1.5874795552202773e-05, | |
| "loss": 0.0752, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 3.4628378378378377, | |
| "grad_norm": 0.857235124255491, | |
| "learning_rate": 1.5660305470741603e-05, | |
| "loss": 0.0819, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.47972972972973, | |
| "grad_norm": 0.9031578400293898, | |
| "learning_rate": 1.5447293153316163e-05, | |
| "loss": 0.0729, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 3.4966216216216215, | |
| "grad_norm": 0.912864587424618, | |
| "learning_rate": 1.523578518692505e-05, | |
| "loss": 0.0835, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 3.5135135135135136, | |
| "grad_norm": 0.8737092915969449, | |
| "learning_rate": 1.5025807970802252e-05, | |
| "loss": 0.0779, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 3.5304054054054053, | |
| "grad_norm": 0.834810088012881, | |
| "learning_rate": 1.481738771312209e-05, | |
| "loss": 0.077, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 3.5472972972972974, | |
| "grad_norm": 0.8458941272972602, | |
| "learning_rate": 1.4610550427728103e-05, | |
| "loss": 0.0776, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.564189189189189, | |
| "grad_norm": 0.8428642842357928, | |
| "learning_rate": 1.4405321930886161e-05, | |
| "loss": 0.0793, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 3.581081081081081, | |
| "grad_norm": 0.7664417243233291, | |
| "learning_rate": 1.4201727838062181e-05, | |
| "loss": 0.0746, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 3.597972972972973, | |
| "grad_norm": 0.8901858386185703, | |
| "learning_rate": 1.3999793560724966e-05, | |
| "loss": 0.0769, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 3.614864864864865, | |
| "grad_norm": 0.8735041757018948, | |
| "learning_rate": 1.3799544303174514e-05, | |
| "loss": 0.0756, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 3.631756756756757, | |
| "grad_norm": 0.7769463844925831, | |
| "learning_rate": 1.3601005059396104e-05, | |
| "loss": 0.074, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.6486486486486487, | |
| "grad_norm": 0.832941867148951, | |
| "learning_rate": 1.3404200609940754e-05, | |
| "loss": 0.0768, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 3.6655405405405403, | |
| "grad_norm": 0.8426081644466242, | |
| "learning_rate": 1.3209155518832203e-05, | |
| "loss": 0.0749, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 3.6824324324324325, | |
| "grad_norm": 0.9273419971193093, | |
| "learning_rate": 1.3015894130500977e-05, | |
| "loss": 0.0728, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 3.6993243243243246, | |
| "grad_norm": 0.8455533228945179, | |
| "learning_rate": 1.2824440566745865e-05, | |
| "loss": 0.0713, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 3.7162162162162162, | |
| "grad_norm": 0.8554325418217714, | |
| "learning_rate": 1.2634818723723174e-05, | |
| "loss": 0.0757, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.733108108108108, | |
| "grad_norm": 0.8201944585960416, | |
| "learning_rate": 1.2447052268964122e-05, | |
| "loss": 0.0692, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.8991946334049652, | |
| "learning_rate": 1.2261164638420832e-05, | |
| "loss": 0.069, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 3.766891891891892, | |
| "grad_norm": 0.825382132844153, | |
| "learning_rate": 1.2077179033541139e-05, | |
| "loss": 0.0712, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 3.7837837837837838, | |
| "grad_norm": 0.793074622490174, | |
| "learning_rate": 1.1895118418372734e-05, | |
| "loss": 0.0753, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 3.8006756756756754, | |
| "grad_norm": 0.7555696916585037, | |
| "learning_rate": 1.171500551669697e-05, | |
| "loss": 0.0694, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.8175675675675675, | |
| "grad_norm": 0.8346510652902012, | |
| "learning_rate": 1.1536862809192518e-05, | |
| "loss": 0.0698, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 3.8344594594594597, | |
| "grad_norm": 0.836330470853335, | |
| "learning_rate": 1.1360712530629513e-05, | |
| "loss": 0.0742, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 3.8513513513513513, | |
| "grad_norm": 0.7869441843004726, | |
| "learning_rate": 1.1186576667094342e-05, | |
| "loss": 0.0709, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 3.868243243243243, | |
| "grad_norm": 0.8183714182060285, | |
| "learning_rate": 1.1014476953245423e-05, | |
| "loss": 0.0692, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 3.885135135135135, | |
| "grad_norm": 0.7866580509216339, | |
| "learning_rate": 1.0844434869600428e-05, | |
| "loss": 0.0697, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.902027027027027, | |
| "grad_norm": 0.8685504391204759, | |
| "learning_rate": 1.0676471639855235e-05, | |
| "loss": 0.0717, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 3.918918918918919, | |
| "grad_norm": 0.7572894210884165, | |
| "learning_rate": 1.0510608228234848e-05, | |
| "loss": 0.0694, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.935810810810811, | |
| "grad_norm": 0.752642827805834, | |
| "learning_rate": 1.0346865336876809e-05, | |
| "loss": 0.0683, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 3.9527027027027026, | |
| "grad_norm": 0.8356745483123115, | |
| "learning_rate": 1.0185263403247256e-05, | |
| "loss": 0.0659, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.9695945945945947, | |
| "grad_norm": 0.8778946453117659, | |
| "learning_rate": 1.002582259759002e-05, | |
| "loss": 0.0679, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.9864864864864864, | |
| "grad_norm": 0.9091930366746289, | |
| "learning_rate": 9.868562820409103e-06, | |
| "loss": 0.0716, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.003378378378378, | |
| "grad_norm": 0.4280982875248494, | |
| "learning_rate": 9.713503699984825e-06, | |
| "loss": 0.0546, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.02027027027027, | |
| "grad_norm": 0.4036282601764634, | |
| "learning_rate": 9.560664589923895e-06, | |
| "loss": 0.0216, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.037162162162162, | |
| "grad_norm": 0.35859643293655324, | |
| "learning_rate": 9.410064566743841e-06, | |
| "loss": 0.0212, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.054054054054054, | |
| "grad_norm": 0.41435811979990983, | |
| "learning_rate": 9.261722427491953e-06, | |
| "loss": 0.0205, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.070945945945946, | |
| "grad_norm": 0.5677279791425531, | |
| "learning_rate": 9.115656687399149e-06, | |
| "loss": 0.0188, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.087837837837838, | |
| "grad_norm": 0.4965394018576192, | |
| "learning_rate": 8.971885577569058e-06, | |
| "loss": 0.0178, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.10472972972973, | |
| "grad_norm": 0.4501724877394163, | |
| "learning_rate": 8.83042704270243e-06, | |
| "loss": 0.0186, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.121621621621622, | |
| "grad_norm": 0.46429914065223105, | |
| "learning_rate": 8.691298738857432e-06, | |
| "loss": 0.0178, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.138513513513513, | |
| "grad_norm": 0.5862559180728094, | |
| "learning_rate": 8.554518031245934e-06, | |
| "loss": 0.0195, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.155405405405405, | |
| "grad_norm": 0.40012503088293433, | |
| "learning_rate": 8.420101992066028e-06, | |
| "loss": 0.0178, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.172297297297297, | |
| "grad_norm": 0.48420732126049076, | |
| "learning_rate": 8.288067398371214e-06, | |
| "loss": 0.0168, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 4.1891891891891895, | |
| "grad_norm": 0.4065785940795942, | |
| "learning_rate": 8.158430729976372e-06, | |
| "loss": 0.0189, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 4.206081081081081, | |
| "grad_norm": 0.4907199226940373, | |
| "learning_rate": 8.031208167400833e-06, | |
| "loss": 0.0184, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 4.222972972972973, | |
| "grad_norm": 0.5139722509677537, | |
| "learning_rate": 7.906415589848834e-06, | |
| "loss": 0.0191, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.239864864864865, | |
| "grad_norm": 0.3932459122852539, | |
| "learning_rate": 7.78406857322756e-06, | |
| "loss": 0.017, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 4.256756756756757, | |
| "grad_norm": 0.4917048473564634, | |
| "learning_rate": 7.664182388203037e-06, | |
| "loss": 0.0192, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 4.273648648648648, | |
| "grad_norm": 0.49328061068221307, | |
| "learning_rate": 7.54677199829414e-06, | |
| "loss": 0.0175, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 4.29054054054054, | |
| "grad_norm": 0.5069962368547102, | |
| "learning_rate": 7.4318520580049444e-06, | |
| "loss": 0.0167, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 4.3074324324324325, | |
| "grad_norm": 0.45503300031661426, | |
| "learning_rate": 7.3194369109956e-06, | |
| "loss": 0.0181, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.324324324324325, | |
| "grad_norm": 0.4688634600711951, | |
| "learning_rate": 7.209540588292083e-06, | |
| "loss": 0.0154, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 4.341216216216216, | |
| "grad_norm": 0.43975517813047615, | |
| "learning_rate": 7.102176806534873e-06, | |
| "loss": 0.0174, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 4.358108108108108, | |
| "grad_norm": 0.49349351945560754, | |
| "learning_rate": 6.9973589662669455e-06, | |
| "loss": 0.0173, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 4.375, | |
| "grad_norm": 0.4598136925028294, | |
| "learning_rate": 6.8951001502612065e-06, | |
| "loss": 0.018, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 4.391891891891892, | |
| "grad_norm": 0.4783402057258077, | |
| "learning_rate": 6.7954131218875404e-06, | |
| "loss": 0.0162, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.408783783783784, | |
| "grad_norm": 0.4185232247836625, | |
| "learning_rate": 6.69831032351977e-06, | |
| "loss": 0.0172, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 4.425675675675675, | |
| "grad_norm": 0.4040507147266634, | |
| "learning_rate": 6.603803874982687e-06, | |
| "loss": 0.0164, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 4.4425675675675675, | |
| "grad_norm": 0.418881018157173, | |
| "learning_rate": 6.511905572039298e-06, | |
| "loss": 0.0183, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 4.45945945945946, | |
| "grad_norm": 0.46713689101304967, | |
| "learning_rate": 6.422626884918559e-06, | |
| "loss": 0.0176, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 4.476351351351352, | |
| "grad_norm": 0.43989858747637034, | |
| "learning_rate": 6.33597895688373e-06, | |
| "loss": 0.0153, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.493243243243243, | |
| "grad_norm": 0.5858107139869251, | |
| "learning_rate": 6.2519726028415145e-06, | |
| "loss": 0.0168, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 4.510135135135135, | |
| "grad_norm": 0.42880372637640857, | |
| "learning_rate": 6.170618307992231e-06, | |
| "loss": 0.0168, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 4.527027027027027, | |
| "grad_norm": 0.43939563736912374, | |
| "learning_rate": 6.091926226521089e-06, | |
| "loss": 0.016, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 4.543918918918919, | |
| "grad_norm": 0.44821397296457444, | |
| "learning_rate": 6.015906180330808e-06, | |
| "loss": 0.0164, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 4.5608108108108105, | |
| "grad_norm": 0.44489436762013035, | |
| "learning_rate": 5.942567657815696e-06, | |
| "loss": 0.0169, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.577702702702703, | |
| "grad_norm": 0.5275823096329626, | |
| "learning_rate": 5.871919812677383e-06, | |
| "loss": 0.0161, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 4.594594594594595, | |
| "grad_norm": 0.4838695631385133, | |
| "learning_rate": 5.8039714627822754e-06, | |
| "loss": 0.0155, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 4.611486486486487, | |
| "grad_norm": 0.45828429488716665, | |
| "learning_rate": 5.738731089060995e-06, | |
| "loss": 0.0147, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 4.628378378378378, | |
| "grad_norm": 0.5039221100761726, | |
| "learning_rate": 5.676206834449797e-06, | |
| "loss": 0.0166, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 4.64527027027027, | |
| "grad_norm": 0.4323629744407566, | |
| "learning_rate": 5.616406502874251e-06, | |
| "loss": 0.0144, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.662162162162162, | |
| "grad_norm": 0.39180413197402647, | |
| "learning_rate": 5.55933755827518e-06, | |
| "loss": 0.015, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 4.679054054054054, | |
| "grad_norm": 0.5188515254318021, | |
| "learning_rate": 5.505007123677063e-06, | |
| "loss": 0.0155, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 4.695945945945946, | |
| "grad_norm": 0.4352611281412726, | |
| "learning_rate": 5.453421980298957e-06, | |
| "loss": 0.0152, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 4.712837837837838, | |
| "grad_norm": 0.44967726061497926, | |
| "learning_rate": 5.4045885667081375e-06, | |
| "loss": 0.0161, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 4.72972972972973, | |
| "grad_norm": 0.4588466562342778, | |
| "learning_rate": 5.358512978016445e-06, | |
| "loss": 0.0156, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.746621621621622, | |
| "grad_norm": 0.5035061702054565, | |
| "learning_rate": 5.315200965119541e-06, | |
| "loss": 0.0162, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 4.763513513513513, | |
| "grad_norm": 0.4333645143853134, | |
| "learning_rate": 5.27465793397911e-06, | |
| "loss": 0.0155, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 4.780405405405405, | |
| "grad_norm": 0.3946393039972133, | |
| "learning_rate": 5.236888944948117e-06, | |
| "loss": 0.0139, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 4.797297297297297, | |
| "grad_norm": 0.5052154455542381, | |
| "learning_rate": 5.201898712139201e-06, | |
| "loss": 0.0152, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 4.8141891891891895, | |
| "grad_norm": 0.4651474482374252, | |
| "learning_rate": 5.1696916028362964e-06, | |
| "loss": 0.0156, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 4.831081081081081, | |
| "grad_norm": 0.48548100702560065, | |
| "learning_rate": 5.1402716369495194e-06, | |
| "loss": 0.016, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 4.847972972972973, | |
| "grad_norm": 0.541954209956823, | |
| "learning_rate": 5.113642486513428e-06, | |
| "loss": 0.0153, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 4.864864864864865, | |
| "grad_norm": 0.4361644908440262, | |
| "learning_rate": 5.089807475228711e-06, | |
| "loss": 0.0136, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 4.881756756756757, | |
| "grad_norm": 0.5665164351885903, | |
| "learning_rate": 5.06876957804733e-06, | |
| "loss": 0.0166, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 4.898648648648649, | |
| "grad_norm": 0.4085257367805478, | |
| "learning_rate": 5.050531420801205e-06, | |
| "loss": 0.0147, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.91554054054054, | |
| "grad_norm": 0.38586369741019666, | |
| "learning_rate": 5.03509527987448e-06, | |
| "loss": 0.0155, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 4.9324324324324325, | |
| "grad_norm": 0.49715228224020613, | |
| "learning_rate": 5.022463081919386e-06, | |
| "loss": 0.0141, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 4.949324324324325, | |
| "grad_norm": 0.37623948634094945, | |
| "learning_rate": 5.012636403615775e-06, | |
| "loss": 0.016, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 4.966216216216216, | |
| "grad_norm": 0.4510174340251159, | |
| "learning_rate": 5.005616471474332e-06, | |
| "loss": 0.0148, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 4.983108108108108, | |
| "grad_norm": 0.4307246592198199, | |
| "learning_rate": 5.001404161683473e-06, | |
| "loss": 0.0145, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.31889541013471795, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0135, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1480, | |
| "total_flos": 688205834289152.0, | |
| "train_loss": 0.36633790682598544, | |
| "train_runtime": 23733.8855, | |
| "train_samples_per_second": 1.993, | |
| "train_steps_per_second": 0.062 | |
| } | |
| ], | |
| "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": 688205834289152.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |