|
{ |
|
"best_metric": 0.4145749807357788, |
|
"best_model_checkpoint": "Action_all_10_class/checkpoint-1400", |
|
"epoch": 5.0, |
|
"eval_steps": 100, |
|
"global_step": 1445, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.03, |
|
"grad_norm": 2.3712973594665527, |
|
"learning_rate": 0.0001986159169550173, |
|
"loss": 2.271, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"grad_norm": 1.8230082988739014, |
|
"learning_rate": 0.0001972318339100346, |
|
"loss": 2.0335, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"grad_norm": 2.621448040008545, |
|
"learning_rate": 0.0001958477508650519, |
|
"loss": 1.8885, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"grad_norm": 3.194303274154663, |
|
"learning_rate": 0.0001944636678200692, |
|
"loss": 1.6835, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"grad_norm": 2.9692325592041016, |
|
"learning_rate": 0.0001930795847750865, |
|
"loss": 1.5658, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"grad_norm": 3.6100592613220215, |
|
"learning_rate": 0.00019169550173010383, |
|
"loss": 1.4881, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"grad_norm": 2.926745653152466, |
|
"learning_rate": 0.00019031141868512113, |
|
"loss": 1.2581, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"grad_norm": 3.4733943939208984, |
|
"learning_rate": 0.00018892733564013843, |
|
"loss": 1.1786, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"grad_norm": 3.609900951385498, |
|
"learning_rate": 0.00018754325259515573, |
|
"loss": 1.1612, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"grad_norm": 3.1664867401123047, |
|
"learning_rate": 0.00018615916955017303, |
|
"loss": 1.1239, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"eval_accuracy": 0.7116564417177914, |
|
"eval_loss": 0.9933902025222778, |
|
"eval_runtime": 13.7579, |
|
"eval_samples_per_second": 59.239, |
|
"eval_steps_per_second": 7.414, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"grad_norm": 3.3514106273651123, |
|
"learning_rate": 0.00018477508650519033, |
|
"loss": 1.1763, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"grad_norm": 2.530064821243286, |
|
"learning_rate": 0.0001833910034602076, |
|
"loss": 1.1493, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"grad_norm": 4.615677833557129, |
|
"learning_rate": 0.0001820069204152249, |
|
"loss": 1.1661, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"grad_norm": 5.2710065841674805, |
|
"learning_rate": 0.0001807612456747405, |
|
"loss": 1.1174, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"grad_norm": 2.854151487350464, |
|
"learning_rate": 0.0001793771626297578, |
|
"loss": 1.0679, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"grad_norm": 4.68364143371582, |
|
"learning_rate": 0.0001779930795847751, |
|
"loss": 1.1566, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"grad_norm": 5.799492359161377, |
|
"learning_rate": 0.0001766089965397924, |
|
"loss": 0.9811, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"grad_norm": 7.8334760665893555, |
|
"learning_rate": 0.0001752249134948097, |
|
"loss": 1.0781, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"grad_norm": 4.817080020904541, |
|
"learning_rate": 0.000173840830449827, |
|
"loss": 1.0302, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"grad_norm": 4.084784507751465, |
|
"learning_rate": 0.0001724567474048443, |
|
"loss": 0.8878, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"eval_accuracy": 0.7705521472392638, |
|
"eval_loss": 0.7667088508605957, |
|
"eval_runtime": 10.9837, |
|
"eval_samples_per_second": 74.201, |
|
"eval_steps_per_second": 9.286, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"grad_norm": 3.8510003089904785, |
|
"learning_rate": 0.0001710726643598616, |
|
"loss": 1.0539, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"grad_norm": 4.4905619621276855, |
|
"learning_rate": 0.00016968858131487892, |
|
"loss": 0.9855, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 3.903388738632202, |
|
"learning_rate": 0.00016830449826989622, |
|
"loss": 1.0442, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"grad_norm": 4.052041530609131, |
|
"learning_rate": 0.00016692041522491352, |
|
"loss": 1.004, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"grad_norm": 4.503437042236328, |
|
"learning_rate": 0.00016553633217993081, |
|
"loss": 0.8825, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"grad_norm": 4.2403459548950195, |
|
"learning_rate": 0.00016415224913494811, |
|
"loss": 0.9616, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"grad_norm": 6.690958499908447, |
|
"learning_rate": 0.00016276816608996541, |
|
"loss": 0.9127, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"grad_norm": 6.591899394989014, |
|
"learning_rate": 0.0001613840830449827, |
|
"loss": 1.0545, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"grad_norm": 2.814940929412842, |
|
"learning_rate": 0.0001601384083044983, |
|
"loss": 0.9937, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"grad_norm": 5.890912055969238, |
|
"learning_rate": 0.00015875432525951557, |
|
"loss": 0.9037, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 1.04, |
|
"eval_accuracy": 0.8098159509202454, |
|
"eval_loss": 0.6369422674179077, |
|
"eval_runtime": 11.1962, |
|
"eval_samples_per_second": 72.793, |
|
"eval_steps_per_second": 9.11, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 1.07, |
|
"grad_norm": 3.311278820037842, |
|
"learning_rate": 0.00015737024221453287, |
|
"loss": 0.825, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 1.11, |
|
"grad_norm": 4.0047454833984375, |
|
"learning_rate": 0.00015598615916955017, |
|
"loss": 0.7347, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 1.14, |
|
"grad_norm": 4.145818710327148, |
|
"learning_rate": 0.00015460207612456747, |
|
"loss": 0.8906, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 1.18, |
|
"grad_norm": 5.543643951416016, |
|
"learning_rate": 0.00015321799307958477, |
|
"loss": 0.7669, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 1.21, |
|
"grad_norm": 5.490930080413818, |
|
"learning_rate": 0.00015183391003460207, |
|
"loss": 0.774, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 1.25, |
|
"grad_norm": 4.895139217376709, |
|
"learning_rate": 0.00015044982698961937, |
|
"loss": 0.6658, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 1.28, |
|
"grad_norm": 4.527533054351807, |
|
"learning_rate": 0.00014906574394463667, |
|
"loss": 0.8603, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 1.31, |
|
"grad_norm": 3.3609440326690674, |
|
"learning_rate": 0.00014768166089965397, |
|
"loss": 0.7099, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 1.35, |
|
"grad_norm": 10.356605529785156, |
|
"learning_rate": 0.0001462975778546713, |
|
"loss": 0.7262, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 1.38, |
|
"grad_norm": 3.6397433280944824, |
|
"learning_rate": 0.0001449134948096886, |
|
"loss": 0.7307, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 1.38, |
|
"eval_accuracy": 0.8319018404907975, |
|
"eval_loss": 0.577174186706543, |
|
"eval_runtime": 11.1343, |
|
"eval_samples_per_second": 73.197, |
|
"eval_steps_per_second": 9.161, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 1.42, |
|
"grad_norm": 5.693297386169434, |
|
"learning_rate": 0.0001435294117647059, |
|
"loss": 0.8061, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 1.45, |
|
"grad_norm": 3.4219260215759277, |
|
"learning_rate": 0.0001421453287197232, |
|
"loss": 0.6227, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 1.49, |
|
"grad_norm": 9.225582122802734, |
|
"learning_rate": 0.0001407612456747405, |
|
"loss": 0.6429, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 1.52, |
|
"grad_norm": 5.8882293701171875, |
|
"learning_rate": 0.0001393771626297578, |
|
"loss": 0.686, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 1.56, |
|
"grad_norm": 3.3229711055755615, |
|
"learning_rate": 0.0001379930795847751, |
|
"loss": 0.7574, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 1.59, |
|
"grad_norm": 5.165830135345459, |
|
"learning_rate": 0.0001366089965397924, |
|
"loss": 0.6195, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 1.63, |
|
"grad_norm": 6.676501274108887, |
|
"learning_rate": 0.0001352249134948097, |
|
"loss": 0.6789, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 1.66, |
|
"grad_norm": 6.1190643310546875, |
|
"learning_rate": 0.000133840830449827, |
|
"loss": 0.7149, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 1.7, |
|
"grad_norm": 2.723904848098755, |
|
"learning_rate": 0.0001324567474048443, |
|
"loss": 0.7014, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 1.73, |
|
"grad_norm": 4.995339870452881, |
|
"learning_rate": 0.00013107266435986162, |
|
"loss": 0.6624, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.73, |
|
"eval_accuracy": 0.7717791411042945, |
|
"eval_loss": 0.6924724578857422, |
|
"eval_runtime": 11.3854, |
|
"eval_samples_per_second": 71.583, |
|
"eval_steps_per_second": 8.959, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.76, |
|
"grad_norm": 5.440706729888916, |
|
"learning_rate": 0.00012968858131487892, |
|
"loss": 0.8681, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 1.8, |
|
"grad_norm": 4.539951801300049, |
|
"learning_rate": 0.0001283044982698962, |
|
"loss": 0.5997, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 1.83, |
|
"grad_norm": 4.5558013916015625, |
|
"learning_rate": 0.0001269204152249135, |
|
"loss": 0.6839, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 1.87, |
|
"grad_norm": 4.361762046813965, |
|
"learning_rate": 0.0001255363321799308, |
|
"loss": 0.6016, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 1.9, |
|
"grad_norm": 4.278672695159912, |
|
"learning_rate": 0.0001241522491349481, |
|
"loss": 0.6111, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 1.94, |
|
"grad_norm": 5.990556240081787, |
|
"learning_rate": 0.0001227681660899654, |
|
"loss": 0.9729, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 1.97, |
|
"grad_norm": 4.511960983276367, |
|
"learning_rate": 0.0001213840830449827, |
|
"loss": 0.6829, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 2.01, |
|
"grad_norm": 3.139665365219116, |
|
"learning_rate": 0.00012, |
|
"loss": 0.6516, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 2.04, |
|
"grad_norm": 2.4773037433624268, |
|
"learning_rate": 0.0001186159169550173, |
|
"loss": 0.492, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 2.08, |
|
"grad_norm": 1.613572597503662, |
|
"learning_rate": 0.0001172318339100346, |
|
"loss": 0.5781, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.08, |
|
"eval_accuracy": 0.8404907975460123, |
|
"eval_loss": 0.5438538789749146, |
|
"eval_runtime": 11.1336, |
|
"eval_samples_per_second": 73.202, |
|
"eval_steps_per_second": 9.161, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.11, |
|
"grad_norm": 2.5122666358947754, |
|
"learning_rate": 0.0001158477508650519, |
|
"loss": 0.55, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 2.15, |
|
"grad_norm": 5.037026882171631, |
|
"learning_rate": 0.00011446366782006921, |
|
"loss": 0.5404, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 2.18, |
|
"grad_norm": 6.795055866241455, |
|
"learning_rate": 0.00011307958477508651, |
|
"loss": 0.6455, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 2.21, |
|
"grad_norm": 6.474560260772705, |
|
"learning_rate": 0.00011169550173010381, |
|
"loss": 0.5184, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 2.25, |
|
"grad_norm": 8.110054969787598, |
|
"learning_rate": 0.00011031141868512111, |
|
"loss": 0.4706, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 2.28, |
|
"grad_norm": 2.1118876934051514, |
|
"learning_rate": 0.00010892733564013841, |
|
"loss": 0.5291, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 2.32, |
|
"grad_norm": 5.386865615844727, |
|
"learning_rate": 0.00010754325259515571, |
|
"loss": 0.605, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 2.35, |
|
"grad_norm": 4.501631736755371, |
|
"learning_rate": 0.00010615916955017303, |
|
"loss": 0.4696, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 2.39, |
|
"grad_norm": 4.06800651550293, |
|
"learning_rate": 0.00010477508650519033, |
|
"loss": 0.5214, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 2.42, |
|
"grad_norm": 3.3574037551879883, |
|
"learning_rate": 0.00010339100346020762, |
|
"loss": 0.5537, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 2.42, |
|
"eval_accuracy": 0.8331288343558282, |
|
"eval_loss": 0.5256926417350769, |
|
"eval_runtime": 11.3468, |
|
"eval_samples_per_second": 71.826, |
|
"eval_steps_per_second": 8.989, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 2.46, |
|
"grad_norm": 6.940188884735107, |
|
"learning_rate": 0.00010200692041522492, |
|
"loss": 0.5371, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 2.49, |
|
"grad_norm": 5.517166614532471, |
|
"learning_rate": 0.00010076124567474047, |
|
"loss": 0.5419, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"grad_norm": 3.2714200019836426, |
|
"learning_rate": 9.937716262975779e-05, |
|
"loss": 0.6657, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 2.56, |
|
"grad_norm": 2.734272003173828, |
|
"learning_rate": 9.79930795847751e-05, |
|
"loss": 0.5712, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 2.6, |
|
"grad_norm": 7.275644302368164, |
|
"learning_rate": 9.66089965397924e-05, |
|
"loss": 0.6513, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 2.63, |
|
"grad_norm": 4.910625457763672, |
|
"learning_rate": 9.52249134948097e-05, |
|
"loss": 0.5641, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 2.66, |
|
"grad_norm": 2.4771077632904053, |
|
"learning_rate": 9.384083044982698e-05, |
|
"loss": 0.4163, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 2.7, |
|
"grad_norm": 4.3278303146362305, |
|
"learning_rate": 9.24567474048443e-05, |
|
"loss": 0.5759, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 2.73, |
|
"grad_norm": 8.122814178466797, |
|
"learning_rate": 9.10726643598616e-05, |
|
"loss": 0.6389, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 2.77, |
|
"grad_norm": 6.369782447814941, |
|
"learning_rate": 8.96885813148789e-05, |
|
"loss": 0.4112, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 2.77, |
|
"eval_accuracy": 0.8564417177914111, |
|
"eval_loss": 0.4499729573726654, |
|
"eval_runtime": 11.5138, |
|
"eval_samples_per_second": 70.785, |
|
"eval_steps_per_second": 8.859, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"grad_norm": 7.130208969116211, |
|
"learning_rate": 8.83044982698962e-05, |
|
"loss": 0.4534, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 2.84, |
|
"grad_norm": 5.094985008239746, |
|
"learning_rate": 8.69204152249135e-05, |
|
"loss": 0.4881, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 2.87, |
|
"grad_norm": 2.527794361114502, |
|
"learning_rate": 8.55363321799308e-05, |
|
"loss": 0.2775, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 2.91, |
|
"grad_norm": 9.366188049316406, |
|
"learning_rate": 8.415224913494811e-05, |
|
"loss": 0.6528, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 2.94, |
|
"grad_norm": 3.863041400909424, |
|
"learning_rate": 8.276816608996541e-05, |
|
"loss": 0.5632, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 2.98, |
|
"grad_norm": 3.3894691467285156, |
|
"learning_rate": 8.138408304498271e-05, |
|
"loss": 0.5885, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 3.01, |
|
"grad_norm": 5.836617469787598, |
|
"learning_rate": 8e-05, |
|
"loss": 0.5864, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 3.04, |
|
"grad_norm": 4.380435943603516, |
|
"learning_rate": 7.86159169550173e-05, |
|
"loss": 0.4523, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 3.08, |
|
"grad_norm": 0.4667840898036957, |
|
"learning_rate": 7.72318339100346e-05, |
|
"loss": 0.2999, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 3.11, |
|
"grad_norm": 3.003122091293335, |
|
"learning_rate": 7.58477508650519e-05, |
|
"loss": 0.3263, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 3.11, |
|
"eval_accuracy": 0.841717791411043, |
|
"eval_loss": 0.4910809397697449, |
|
"eval_runtime": 11.4121, |
|
"eval_samples_per_second": 71.416, |
|
"eval_steps_per_second": 8.938, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 3.15, |
|
"grad_norm": 4.519166469573975, |
|
"learning_rate": 7.44636678200692e-05, |
|
"loss": 0.417, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 3.18, |
|
"grad_norm": 6.071402549743652, |
|
"learning_rate": 7.30795847750865e-05, |
|
"loss": 0.351, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 3.22, |
|
"grad_norm": 6.0896687507629395, |
|
"learning_rate": 7.16955017301038e-05, |
|
"loss": 0.4687, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 3.25, |
|
"grad_norm": 4.001879692077637, |
|
"learning_rate": 7.031141868512112e-05, |
|
"loss": 0.5083, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 3.29, |
|
"grad_norm": 5.064093112945557, |
|
"learning_rate": 6.892733564013842e-05, |
|
"loss": 0.3407, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 3.32, |
|
"grad_norm": 1.3715455532073975, |
|
"learning_rate": 6.754325259515572e-05, |
|
"loss": 0.4073, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 3.36, |
|
"grad_norm": 6.380220413208008, |
|
"learning_rate": 6.615916955017302e-05, |
|
"loss": 0.4613, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 3.39, |
|
"grad_norm": 2.1010074615478516, |
|
"learning_rate": 6.477508650519032e-05, |
|
"loss": 0.3706, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 3.43, |
|
"grad_norm": 2.2640676498413086, |
|
"learning_rate": 6.339100346020761e-05, |
|
"loss": 0.3235, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 3.46, |
|
"grad_norm": 2.6152210235595703, |
|
"learning_rate": 6.200692041522491e-05, |
|
"loss": 0.4592, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 3.46, |
|
"eval_accuracy": 0.8711656441717791, |
|
"eval_loss": 0.4550653100013733, |
|
"eval_runtime": 11.3924, |
|
"eval_samples_per_second": 71.539, |
|
"eval_steps_per_second": 8.953, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 3.49, |
|
"grad_norm": 4.868244647979736, |
|
"learning_rate": 6.0622837370242214e-05, |
|
"loss": 0.3891, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 3.53, |
|
"grad_norm": 4.125463962554932, |
|
"learning_rate": 5.9238754325259514e-05, |
|
"loss": 0.3462, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 3.56, |
|
"grad_norm": 5.510716915130615, |
|
"learning_rate": 5.785467128027682e-05, |
|
"loss": 0.3938, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"grad_norm": 4.867416858673096, |
|
"learning_rate": 5.647058823529412e-05, |
|
"loss": 0.3851, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 3.63, |
|
"grad_norm": 5.599556922912598, |
|
"learning_rate": 5.508650519031142e-05, |
|
"loss": 0.3939, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 3.67, |
|
"grad_norm": 6.144674777984619, |
|
"learning_rate": 5.3702422145328725e-05, |
|
"loss": 0.5787, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 3.7, |
|
"grad_norm": 6.78063440322876, |
|
"learning_rate": 5.2318339100346025e-05, |
|
"loss": 0.4548, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 3.74, |
|
"grad_norm": 2.213515281677246, |
|
"learning_rate": 5.0934256055363325e-05, |
|
"loss": 0.418, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 3.77, |
|
"grad_norm": 3.3876266479492188, |
|
"learning_rate": 4.9550173010380624e-05, |
|
"loss": 0.432, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 3.81, |
|
"grad_norm": 6.500645160675049, |
|
"learning_rate": 4.8166089965397924e-05, |
|
"loss": 0.3204, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 3.81, |
|
"eval_accuracy": 0.8723926380368098, |
|
"eval_loss": 0.4324829876422882, |
|
"eval_runtime": 11.2965, |
|
"eval_samples_per_second": 72.146, |
|
"eval_steps_per_second": 9.029, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 3.84, |
|
"grad_norm": 2.518057107925415, |
|
"learning_rate": 4.678200692041523e-05, |
|
"loss": 0.3505, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 3.88, |
|
"grad_norm": 4.820311546325684, |
|
"learning_rate": 4.539792387543253e-05, |
|
"loss": 0.4641, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 3.91, |
|
"grad_norm": 1.1859557628631592, |
|
"learning_rate": 4.401384083044983e-05, |
|
"loss": 0.5202, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 3.94, |
|
"grad_norm": 2.3869946002960205, |
|
"learning_rate": 4.262975778546713e-05, |
|
"loss": 0.5007, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 3.98, |
|
"grad_norm": 6.64663553237915, |
|
"learning_rate": 4.124567474048443e-05, |
|
"loss": 0.3478, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 4.01, |
|
"grad_norm": 4.931898593902588, |
|
"learning_rate": 3.9861591695501735e-05, |
|
"loss": 0.3808, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 4.05, |
|
"grad_norm": 1.7521384954452515, |
|
"learning_rate": 3.8477508650519034e-05, |
|
"loss": 0.3247, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 4.08, |
|
"grad_norm": 1.5513068437576294, |
|
"learning_rate": 3.7093425605536334e-05, |
|
"loss": 0.393, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 4.12, |
|
"grad_norm": 5.2524094581604, |
|
"learning_rate": 3.570934256055363e-05, |
|
"loss": 0.3697, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 4.15, |
|
"grad_norm": 5.346503734588623, |
|
"learning_rate": 3.432525951557093e-05, |
|
"loss": 0.3268, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 4.15, |
|
"eval_accuracy": 0.8539877300613496, |
|
"eval_loss": 0.4529338777065277, |
|
"eval_runtime": 11.243, |
|
"eval_samples_per_second": 72.49, |
|
"eval_steps_per_second": 9.072, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 4.19, |
|
"grad_norm": 2.0786306858062744, |
|
"learning_rate": 3.294117647058824e-05, |
|
"loss": 0.2967, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 4.22, |
|
"grad_norm": 3.8141846656799316, |
|
"learning_rate": 3.155709342560554e-05, |
|
"loss": 0.2483, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 4.26, |
|
"grad_norm": 1.8864765167236328, |
|
"learning_rate": 3.0173010380622842e-05, |
|
"loss": 0.2443, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 4.29, |
|
"grad_norm": 0.9406078457832336, |
|
"learning_rate": 2.878892733564014e-05, |
|
"loss": 0.2637, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 4.33, |
|
"grad_norm": 3.418455123901367, |
|
"learning_rate": 2.7404844290657437e-05, |
|
"loss": 0.2866, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 4.36, |
|
"grad_norm": 1.8662397861480713, |
|
"learning_rate": 2.602076124567474e-05, |
|
"loss": 0.3381, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 4.39, |
|
"grad_norm": 3.9735002517700195, |
|
"learning_rate": 2.4636678200692043e-05, |
|
"loss": 0.3328, |
|
"step": 1270 |
|
}, |
|
{ |
|
"epoch": 4.43, |
|
"grad_norm": 2.229581117630005, |
|
"learning_rate": 2.3252595155709346e-05, |
|
"loss": 0.4579, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 4.46, |
|
"grad_norm": 3.8186144828796387, |
|
"learning_rate": 2.1868512110726642e-05, |
|
"loss": 0.2986, |
|
"step": 1290 |
|
}, |
|
{ |
|
"epoch": 4.5, |
|
"grad_norm": 2.310910701751709, |
|
"learning_rate": 2.0484429065743945e-05, |
|
"loss": 0.4267, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 4.5, |
|
"eval_accuracy": 0.8723926380368098, |
|
"eval_loss": 0.4355594515800476, |
|
"eval_runtime": 11.2323, |
|
"eval_samples_per_second": 72.559, |
|
"eval_steps_per_second": 9.081, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 4.53, |
|
"grad_norm": 5.4986042976379395, |
|
"learning_rate": 1.910034602076125e-05, |
|
"loss": 0.3289, |
|
"step": 1310 |
|
}, |
|
{ |
|
"epoch": 4.57, |
|
"grad_norm": 9.17880916595459, |
|
"learning_rate": 1.7716262975778548e-05, |
|
"loss": 0.4384, |
|
"step": 1320 |
|
}, |
|
{ |
|
"epoch": 4.6, |
|
"grad_norm": 0.5825958251953125, |
|
"learning_rate": 1.6332179930795848e-05, |
|
"loss": 0.1629, |
|
"step": 1330 |
|
}, |
|
{ |
|
"epoch": 4.64, |
|
"grad_norm": 3.7861948013305664, |
|
"learning_rate": 1.4948096885813149e-05, |
|
"loss": 0.2607, |
|
"step": 1340 |
|
}, |
|
{ |
|
"epoch": 4.67, |
|
"grad_norm": 0.08592710644006729, |
|
"learning_rate": 1.356401384083045e-05, |
|
"loss": 0.23, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 4.71, |
|
"grad_norm": 3.897566556930542, |
|
"learning_rate": 1.2179930795847751e-05, |
|
"loss": 0.2927, |
|
"step": 1360 |
|
}, |
|
{ |
|
"epoch": 4.74, |
|
"grad_norm": 3.4900457859039307, |
|
"learning_rate": 1.0795847750865053e-05, |
|
"loss": 0.3149, |
|
"step": 1370 |
|
}, |
|
{ |
|
"epoch": 4.78, |
|
"grad_norm": 0.48032164573669434, |
|
"learning_rate": 9.411764705882354e-06, |
|
"loss": 0.3346, |
|
"step": 1380 |
|
}, |
|
{ |
|
"epoch": 4.81, |
|
"grad_norm": 3.148484945297241, |
|
"learning_rate": 8.027681660899653e-06, |
|
"loss": 0.3006, |
|
"step": 1390 |
|
}, |
|
{ |
|
"epoch": 4.84, |
|
"grad_norm": 8.376299858093262, |
|
"learning_rate": 6.6435986159169555e-06, |
|
"loss": 0.2886, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 4.84, |
|
"eval_accuracy": 0.8785276073619632, |
|
"eval_loss": 0.4145749807357788, |
|
"eval_runtime": 11.483, |
|
"eval_samples_per_second": 70.974, |
|
"eval_steps_per_second": 8.883, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 4.88, |
|
"grad_norm": 10.592811584472656, |
|
"learning_rate": 5.259515570934257e-06, |
|
"loss": 0.2866, |
|
"step": 1410 |
|
}, |
|
{ |
|
"epoch": 4.91, |
|
"grad_norm": 1.9005374908447266, |
|
"learning_rate": 3.875432525951557e-06, |
|
"loss": 0.367, |
|
"step": 1420 |
|
}, |
|
{ |
|
"epoch": 4.95, |
|
"grad_norm": 6.674309730529785, |
|
"learning_rate": 2.4913494809688584e-06, |
|
"loss": 0.2715, |
|
"step": 1430 |
|
}, |
|
{ |
|
"epoch": 4.98, |
|
"grad_norm": 7.622613430023193, |
|
"learning_rate": 1.1072664359861592e-06, |
|
"loss": 0.3306, |
|
"step": 1440 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"step": 1445, |
|
"total_flos": 1.789030847196795e+18, |
|
"train_loss": 0.6435174308433664, |
|
"train_runtime": 776.1745, |
|
"train_samples_per_second": 29.742, |
|
"train_steps_per_second": 1.862 |
|
} |
|
], |
|
"logging_steps": 10, |
|
"max_steps": 1445, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 5, |
|
"save_steps": 100, |
|
"total_flos": 1.789030847196795e+18, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|