| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.9984901862103674, | |
| "global_step": 1240, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.2e-05, | |
| "loss": 6.4085, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.4e-05, | |
| "loss": 5.2039, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.6e-05, | |
| "loss": 4.8094, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.8e-05, | |
| "loss": 4.3557, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 6e-05, | |
| "loss": 4.0778, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 5.999749289566312e-05, | |
| "loss": 3.9019, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 5.998997200169063e-05, | |
| "loss": 3.744, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 5.9977438575126904e-05, | |
| "loss": 3.5981, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 5.99598947108125e-05, | |
| "loss": 3.4746, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 5.993734334103397e-05, | |
| "loss": 3.3712, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5.990978823503376e-05, | |
| "loss": 3.3405, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 5.987723399838028e-05, | |
| "loss": 3.3019, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5.983968607219804e-05, | |
| "loss": 3.2548, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 5.979715073225829e-05, | |
| "loss": 3.2847, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5.974963508793003e-05, | |
| "loss": 3.2371, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 5.9697147080991817e-05, | |
| "loss": 3.1677, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 5.963969548430428e-05, | |
| "loss": 3.1617, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 5.957728990034389e-05, | |
| "loss": 3.2266, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 5.950994075959803e-05, | |
| "loss": 3.1008, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5.9437659318821544e-05, | |
| "loss": 3.0858, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 5.9360457659155325e-05, | |
| "loss": 3.018, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5.927834868410711e-05, | |
| "loss": 3.0951, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5.919134611739472e-05, | |
| "loss": 3.1038, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5.90994645006523e-05, | |
| "loss": 3.0275, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5.900271919099987e-05, | |
| "loss": 3.0632, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5.8901126358476414e-05, | |
| "loss": 3.0435, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5.8794702983337354e-05, | |
| "loss": 3.0403, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5.868346685321639e-05, | |
| "loss": 3.0174, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 5.856743656015246e-05, | |
| "loss": 2.9917, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 5.844663149748229e-05, | |
| "loss": 2.9663, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 5.8321071856599e-05, | |
| "loss": 2.8901, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 5.819077862357725e-05, | |
| "loss": 2.9219, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 5.8055773575665695e-05, | |
| "loss": 2.9779, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 5.791607927764706e-05, | |
| "loss": 2.8977, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 5.777171907806672e-05, | |
| "loss": 2.8961, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 5.762271710533015e-05, | |
| "loss": 2.8841, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5.746909826367018e-05, | |
| "loss": 2.8661, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 5.73108882289844e-05, | |
| "loss": 2.8671, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5.714811344454376e-05, | |
| "loss": 2.9775, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.698080111657279e-05, | |
| "loss": 2.9726, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 5.680897920970237e-05, | |
| "loss": 2.9081, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5.663267644229568e-05, | |
| "loss": 2.8349, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 5.645192228164824e-05, | |
| "loss": 2.9355, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5.6266746939062725e-05, | |
| "loss": 2.904, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5.607718136479943e-05, | |
| "loss": 2.9315, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5.5883257242903255e-05, | |
| "loss": 2.8953, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 5.568500698590799e-05, | |
| "loss": 2.818, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 5.548246372941892e-05, | |
| "loss": 2.8779, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 5.5275661326574485e-05, | |
| "loss": 2.8299, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.7971479892730713, | |
| "eval_runtime": 163.2621, | |
| "eval_samples_per_second": 49.638, | |
| "eval_steps_per_second": 6.205, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 5.50646343423881e-05, | |
| "loss": 3.1054, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 5.48494180479709e-05, | |
| "loss": 2.8559, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 5.463004841463656e-05, | |
| "loss": 2.8828, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 5.4406562107889024e-05, | |
| "loss": 2.8264, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 5.417899648129423e-05, | |
| "loss": 2.8157, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 5.3947389570236784e-05, | |
| "loss": 2.8592, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 5.371178008556278e-05, | |
| "loss": 2.8296, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 5.347220740710959e-05, | |
| "loss": 2.8519, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 5.322871157712397e-05, | |
| "loss": 2.8342, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 5.298133329356934e-05, | |
| "loss": 2.8232, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 5.273011390332354e-05, | |
| "loss": 2.7835, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 5.247509539526808e-05, | |
| "loss": 2.8445, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 5.221632039327014e-05, | |
| "loss": 2.8093, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 5.195383214905837e-05, | |
| "loss": 2.8389, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 5.168767453499379e-05, | |
| "loss": 2.8264, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 5.141789203673698e-05, | |
| "loss": 2.7895, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 5.114452974581269e-05, | |
| "loss": 2.7584, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 5.0867633352073254e-05, | |
| "loss": 2.7946, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 5.0587249136062016e-05, | |
| "loss": 2.7828, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 5.0303423961277864e-05, | |
| "loss": 2.833, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 5.001620526634258e-05, | |
| "loss": 2.7956, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 4.972564105707191e-05, | |
| "loss": 2.8566, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 4.943177989845176e-05, | |
| "loss": 2.7128, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 4.9134670906521176e-05, | |
| "loss": 2.7443, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 4.883436374016296e-05, | |
| "loss": 2.8008, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 4.8530908592803714e-05, | |
| "loss": 2.7743, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 4.8224356184024506e-05, | |
| "loss": 2.7487, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 4.791475775108359e-05, | |
| "loss": 2.7707, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 4.760216504035254e-05, | |
| "loss": 2.7924, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 4.7286630298667406e-05, | |
| "loss": 2.8114, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 4.6968206264596164e-05, | |
| "loss": 2.7762, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 4.664694615962391e-05, | |
| "loss": 2.7792, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 4.632290367925748e-05, | |
| "loss": 2.8243, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 4.5996132984050744e-05, | |
| "loss": 2.7644, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 4.566668869055216e-05, | |
| "loss": 2.7652, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.5334625862176195e-05, | |
| "loss": 2.7722, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.7455, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 4.4662867033486924e-05, | |
| "loss": 2.7365, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 4.4323283311138474e-05, | |
| "loss": 2.7176, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.398130559107617e-05, | |
| "loss": 2.7344, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 4.363699103155502e-05, | |
| "loss": 2.7425, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 4.3290397181410056e-05, | |
| "loss": 2.7304, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 4.294158197043761e-05, | |
| "loss": 2.7893, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 4.259060369971287e-05, | |
| "loss": 2.7134, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 4.2237521031845506e-05, | |
| "loss": 2.82, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 4.1882392981174704e-05, | |
| "loss": 2.7194, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 4.152527890390553e-05, | |
| "loss": 2.7565, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 4.11662384881881e-05, | |
| "loss": 2.7037, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 4.080533174414131e-05, | |
| "loss": 2.7439, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 4.0442618993822714e-05, | |
| "loss": 2.6984, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.6826047897338867, | |
| "eval_runtime": 163.0962, | |
| "eval_samples_per_second": 49.688, | |
| "eval_steps_per_second": 6.211, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 4.007816086114627e-05, | |
| "loss": 2.8637, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 3.971201826174964e-05, | |
| "loss": 2.709, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 3.934425239281274e-05, | |
| "loss": 2.6646, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 3.897492472282926e-05, | |
| "loss": 2.7183, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 3.860409698133271e-05, | |
| "loss": 2.7506, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 3.8231831148579046e-05, | |
| "loss": 2.7579, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 3.785818944518718e-05, | |
| "loss": 2.685, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 3.748323432173944e-05, | |
| "loss": 2.6694, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 3.710702844834357e-05, | |
| "loss": 2.7386, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 3.672963470415803e-05, | |
| "loss": 2.692, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 3.6351116166882325e-05, | |
| "loss": 2.7219, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 3.5971536102214215e-05, | |
| "loss": 2.769, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 3.559095795327547e-05, | |
| "loss": 2.7076, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 3.520944533000792e-05, | |
| "loss": 2.7279, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 3.482706199854172e-05, | |
| "loss": 2.7273, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 3.444387187053747e-05, | |
| "loss": 2.6949, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 3.405993899250394e-05, | |
| "loss": 2.681, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 3.3675327535093366e-05, | |
| "loss": 2.8151, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 3.329010178237595e-05, | |
| "loss": 2.7137, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 3.2904326121095375e-05, | |
| "loss": 2.6656, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 3.25180650299072e-05, | |
| "loss": 2.7158, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 3.2131383068601905e-05, | |
| "loss": 2.7081, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 3.174434486731428e-05, | |
| "loss": 2.644, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 3.135701511572119e-05, | |
| "loss": 2.698, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 3.0969458552229324e-05, | |
| "loss": 2.6591, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 3.058173995315473e-05, | |
| "loss": 2.6591, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 3.019392412189618e-05, | |
| "loss": 2.7327, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 2.9806075878103817e-05, | |
| "loss": 2.6711, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 2.9418260046845272e-05, | |
| "loss": 2.7034, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 2.9030541447770684e-05, | |
| "loss": 2.6119, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 2.8642984884278812e-05, | |
| "loss": 2.641, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 2.8255655132685727e-05, | |
| "loss": 2.6479, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 2.7868616931398106e-05, | |
| "loss": 2.6298, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 2.7481934970092792e-05, | |
| "loss": 2.6951, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 2.709567387890463e-05, | |
| "loss": 2.681, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 2.6709898217624053e-05, | |
| "loss": 2.6787, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 2.6324672464906643e-05, | |
| "loss": 2.7171, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 2.5940061007496078e-05, | |
| "loss": 2.7116, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 2.5556128129462543e-05, | |
| "loss": 2.6834, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2.517293800145828e-05, | |
| "loss": 2.6827, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 2.479055466999209e-05, | |
| "loss": 2.6158, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2.4409042046724537e-05, | |
| "loss": 2.7102, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 2.4028463897785787e-05, | |
| "loss": 2.6505, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 2.3648883833117687e-05, | |
| "loss": 2.6876, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 2.327036529584198e-05, | |
| "loss": 2.6841, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 2.2892971551656424e-05, | |
| "loss": 2.683, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 2.2516765678260564e-05, | |
| "loss": 2.714, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 2.2141810554812824e-05, | |
| "loss": 2.6786, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "learning_rate": 2.176816885142096e-05, | |
| "loss": 2.7022, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.6361331939697266, | |
| "eval_runtime": 163.5049, | |
| "eval_samples_per_second": 49.564, | |
| "eval_steps_per_second": 6.196, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 2.132162043617224e-05, | |
| "loss": 2.8742, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 2.0951087758432035e-05, | |
| "loss": 2.6404, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "learning_rate": 2.0582067518500164e-05, | |
| "loss": 2.7206, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 2.0214621394526204e-05, | |
| "loss": 2.6592, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "learning_rate": 1.984881080156157e-05, | |
| "loss": 2.6667, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "learning_rate": 1.9484696881294518e-05, | |
| "loss": 2.6568, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 1.9122340491830983e-05, | |
| "loss": 2.645, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "learning_rate": 1.8761802197522638e-05, | |
| "loss": 2.5945, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 1.8403142258844258e-05, | |
| "loss": 2.6536, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 1.8046420622321678e-05, | |
| "loss": 2.6664, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 1.769169691051236e-05, | |
| "loss": 2.6797, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 1.733903041204006e-05, | |
| "loss": 2.6706, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 1.6988480071685293e-05, | |
| "loss": 2.7273, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 1.664010448053332e-05, | |
| "loss": 2.6833, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 1.6293961866181154e-05, | |
| "loss": 2.7188, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "learning_rate": 1.5950110083005433e-05, | |
| "loss": 2.6276, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 1.5608606602492622e-05, | |
| "loss": 2.7026, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "learning_rate": 1.526950850363318e-05, | |
| "loss": 2.6412, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 1.4932872463381392e-05, | |
| "loss": 2.6308, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 1.4598754747182377e-05, | |
| "loss": 2.5865, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 1.4267211199567806e-05, | |
| "loss": 2.7103, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "learning_rate": 1.3938297234822089e-05, | |
| "loss": 2.6321, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "learning_rate": 1.361206782772039e-05, | |
| "loss": 2.6398, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "learning_rate": 1.328857750434012e-05, | |
| "loss": 2.6513, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "learning_rate": 1.2967880332947474e-05, | |
| "loss": 2.692, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 1.2650029914960406e-05, | |
| "loss": 2.5955, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "learning_rate": 1.2335079375989666e-05, | |
| "loss": 2.6501, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "learning_rate": 1.2023081356959398e-05, | |
| "loss": 2.646, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 1.1714088005308714e-05, | |
| "loss": 2.6988, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "learning_rate": 1.140815096627573e-05, | |
| "loss": 2.6395, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 1.1105321374265619e-05, | |
| "loss": 2.7236, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "learning_rate": 1.0805649844303954e-05, | |
| "loss": 2.7175, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 1.0509186463576888e-05, | |
| "loss": 2.7241, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 1.0215980783059585e-05, | |
| "loss": 2.6334, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "learning_rate": 9.926081809234262e-06, | |
| "loss": 2.6283, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 9.639537995899214e-06, | |
| "loss": 2.6598, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 9.356397236070267e-06, | |
| "loss": 2.6099, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 9.076706853975862e-06, | |
| "loss": 2.613, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 8.800513597147349e-06, | |
| "loss": 2.5609, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "learning_rate": 8.527863628605514e-06, | |
| "loss": 2.7093, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "learning_rate": 8.258802519144935e-06, | |
| "loss": 2.5844, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 7.993375239717228e-06, | |
| "loss": 2.6648, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 7.731626153914636e-06, | |
| "loss": 2.6364, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 7.473599010555025e-06, | |
| "loss": 2.738, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "learning_rate": 7.219336936369749e-06, | |
| "loss": 2.6826, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 6.968882428795393e-06, | |
| "loss": 2.717, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 6.722277348870772e-06, | |
| "loss": 2.7357, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "learning_rate": 6.47956291424023e-06, | |
| "loss": 2.648, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "learning_rate": 6.240779692264567e-06, | |
| "loss": 2.6178, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "learning_rate": 6.005967593240522e-06, | |
| "loss": 2.6436, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.6245498657226562, | |
| "eval_runtime": 163.5799, | |
| "eval_samples_per_second": 49.542, | |
| "eval_steps_per_second": 6.193, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "learning_rate": 5.7751658637302344e-06, | |
| "loss": 2.8379, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "learning_rate": 5.548413080001497e-06, | |
| "loss": 2.6199, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 5.325747141580131e-06, | |
| "loss": 2.6231, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 5.107205264915482e-06, | |
| "loss": 2.6267, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "learning_rate": 4.8928239771599935e-06, | |
| "loss": 2.69, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 4.6826391100640785e-06, | |
| "loss": 2.7156, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "learning_rate": 4.476685793987202e-06, | |
| "loss": 2.729, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "learning_rate": 4.274998452026159e-06, | |
| "loss": 2.6732, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 4.0776107942615855e-06, | |
| "loss": 2.6484, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 4.19, | |
| "learning_rate": 3.884555812123685e-06, | |
| "loss": 2.6133, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "learning_rate": 3.6958657728779856e-06, | |
| "loss": 2.6863, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "learning_rate": 3.5115722142321906e-06, | |
| "loss": 2.6385, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "learning_rate": 3.3317059390649906e-06, | |
| "loss": 2.6193, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 3.1562970102776177e-06, | |
| "loss": 2.655, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "learning_rate": 2.985374745769134e-06, | |
| "loss": 2.6629, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "learning_rate": 2.8189677135362635e-06, | |
| "loss": 2.6214, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 2.6571037268984742e-06, | |
| "loss": 2.7275, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 2.499809839849292e-06, | |
| "loss": 2.6338, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "learning_rate": 2.3471123425344863e-06, | |
| "loss": 2.6312, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 2.1990367568578982e-06, | |
| "loss": 2.5713, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 2.055607832215727e-06, | |
| "loss": 2.708, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "learning_rate": 1.9168495413599183e-06, | |
| "loss": 2.6313, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "learning_rate": 1.7827850763913167e-06, | |
| "loss": 2.5938, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 1.6534368448833581e-06, | |
| "loss": 2.6137, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 1.5288264661368523e-06, | |
| "loss": 2.6919, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "learning_rate": 1.4089747675665387e-06, | |
| "loss": 2.6168, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "learning_rate": 1.2939017812199584e-06, | |
| "loss": 2.6687, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 4.56, | |
| "learning_rate": 1.1836267404293311e-06, | |
| "loss": 2.5993, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "learning_rate": 1.0781680765968515e-06, | |
| "loss": 2.6695, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 9.775434161140861e-07, | |
| "loss": 2.6461, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "learning_rate": 8.817695774158729e-07, | |
| "loss": 2.6796, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 7.908625681693138e-07, | |
| "loss": 2.6503, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "learning_rate": 7.048375825982112e-07, | |
| "loss": 2.6532, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "learning_rate": 6.237089989435263e-07, | |
| "loss": 2.6171, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "learning_rate": 5.474903770601913e-07, | |
| "loss": 2.6785, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 4.7619445615070256e-07, | |
| "loss": 2.7045, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "learning_rate": 4.0983315263589004e-07, | |
| "loss": 2.7024, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 4.76, | |
| "learning_rate": 3.484175581632143e-07, | |
| "loss": 2.648, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 4.78, | |
| "learning_rate": 2.9195793775288916e-07, | |
| "loss": 2.6865, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 2.4046372808220084e-07, | |
| "loss": 2.6769, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "learning_rate": 1.939435359082331e-07, | |
| "loss": 2.6381, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "learning_rate": 1.524051366293644e-07, | |
| "loss": 2.6804, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "learning_rate": 1.1585547298565091e-07, | |
| "loss": 2.5962, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 8.43006538984481e-08, | |
| "loss": 2.6744, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 5.774595344934075e-08, | |
| "loss": 2.6654, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 4.92, | |
| "learning_rate": 3.6195809998638144e-08, | |
| "loss": 2.673, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "learning_rate": 1.9653825443547435e-08, | |
| "loss": 2.6151, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "learning_rate": 8.122764616148581e-09, | |
| "loss": 2.7565, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "learning_rate": 1.6045548212872963e-09, | |
| "loss": 2.6162, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 1.0028551442609057e-10, | |
| "loss": 2.6195, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.62469482421875, | |
| "eval_runtime": 163.7545, | |
| "eval_samples_per_second": 49.489, | |
| "eval_steps_per_second": 6.186, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 1240, | |
| "total_flos": 8.30390621259694e+16, | |
| "train_loss": 2.8163739796607725, | |
| "train_runtime": 27047.4241, | |
| "train_samples_per_second": 11.753, | |
| "train_steps_per_second": 0.046 | |
| } | |
| ], | |
| "max_steps": 1240, | |
| "num_train_epochs": 5, | |
| "total_flos": 8.30390621259694e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |