| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 8668, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.046146746654360866, | |
| "grad_norm": 0.21824534237384796, | |
| "learning_rate": 2.23760092272203e-06, | |
| "loss": 3.136, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09229349330872173, | |
| "grad_norm": 0.2850794494152069, | |
| "learning_rate": 4.544405997693195e-06, | |
| "loss": 3.1073, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1384402399630826, | |
| "grad_norm": 0.36149144172668457, | |
| "learning_rate": 6.828143021914648e-06, | |
| "loss": 3.1035, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.18458698661744347, | |
| "grad_norm": 0.5060057640075684, | |
| "learning_rate": 9.134948096885815e-06, | |
| "loss": 3.106, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23073373327180433, | |
| "grad_norm": 0.3851727247238159, | |
| "learning_rate": 1.144175317185698e-05, | |
| "loss": 3.0204, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2768804799261652, | |
| "grad_norm": 0.30784064531326294, | |
| "learning_rate": 1.3748558246828143e-05, | |
| "loss": 2.9708, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.3230272265805261, | |
| "grad_norm": 0.6206746697425842, | |
| "learning_rate": 1.605536332179931e-05, | |
| "loss": 2.8294, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.36917397323488693, | |
| "grad_norm": 0.47670331597328186, | |
| "learning_rate": 1.8362168396770474e-05, | |
| "loss": 2.6824, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4153207198892478, | |
| "grad_norm": 0.4056473672389984, | |
| "learning_rate": 1.9925650557620818e-05, | |
| "loss": 2.4481, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.46146746654360865, | |
| "grad_norm": 0.8275523781776428, | |
| "learning_rate": 1.96692731701064e-05, | |
| "loss": 2.0463, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5076142131979695, | |
| "grad_norm": 0.5123036503791809, | |
| "learning_rate": 1.9412895782591976e-05, | |
| "loss": 1.9237, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.5537609598523304, | |
| "grad_norm": 0.7184427976608276, | |
| "learning_rate": 1.9156518395077554e-05, | |
| "loss": 1.881, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5999077065066912, | |
| "grad_norm": 1.0534940958023071, | |
| "learning_rate": 1.8900141007563134e-05, | |
| "loss": 1.8219, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.6460544531610521, | |
| "grad_norm": 0.9063017964363098, | |
| "learning_rate": 1.8646327393923858e-05, | |
| "loss": 1.7868, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.6922011998154131, | |
| "grad_norm": 0.7229586839675903, | |
| "learning_rate": 1.8389950006409436e-05, | |
| "loss": 1.7581, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7383479464697739, | |
| "grad_norm": 0.6040017604827881, | |
| "learning_rate": 1.8133572618895013e-05, | |
| "loss": 1.7096, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.7844946931241348, | |
| "grad_norm": 0.7014355659484863, | |
| "learning_rate": 1.7877195231380594e-05, | |
| "loss": 1.6372, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.8306414397784956, | |
| "grad_norm": 0.45642441511154175, | |
| "learning_rate": 1.762081784386617e-05, | |
| "loss": 1.5638, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8767881864328565, | |
| "grad_norm": 0.7556698322296143, | |
| "learning_rate": 1.7364440456351752e-05, | |
| "loss": 1.5481, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.9229349330872173, | |
| "grad_norm": 0.48637843132019043, | |
| "learning_rate": 1.710806306883733e-05, | |
| "loss": 1.4976, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.9690816797415782, | |
| "grad_norm": 0.6813339591026306, | |
| "learning_rate": 1.685168568132291e-05, | |
| "loss": 1.4417, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.015228426395939, | |
| "grad_norm": 1.094603180885315, | |
| "learning_rate": 1.6595308293808488e-05, | |
| "loss": 1.3938, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.0613751730503, | |
| "grad_norm": 0.37147483229637146, | |
| "learning_rate": 1.6338930906294065e-05, | |
| "loss": 1.3497, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.1075219197046609, | |
| "grad_norm": 0.2780097424983978, | |
| "learning_rate": 1.6082553518779646e-05, | |
| "loss": 1.3099, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.1536686663590217, | |
| "grad_norm": 0.271342933177948, | |
| "learning_rate": 1.5826176131265223e-05, | |
| "loss": 1.285, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.1998154130133827, | |
| "grad_norm": 0.27299413084983826, | |
| "learning_rate": 1.55697987437508e-05, | |
| "loss": 1.2614, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.2459621596677435, | |
| "grad_norm": 0.29231297969818115, | |
| "learning_rate": 1.531342135623638e-05, | |
| "loss": 1.2308, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.2921089063221043, | |
| "grad_norm": 0.22232797741889954, | |
| "learning_rate": 1.505704396872196e-05, | |
| "loss": 1.1944, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.338255652976465, | |
| "grad_norm": 0.9203324913978577, | |
| "learning_rate": 1.480066658120754e-05, | |
| "loss": 1.1544, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.384402399630826, | |
| "grad_norm": 0.19580288231372833, | |
| "learning_rate": 1.4544289193693117e-05, | |
| "loss": 1.0996, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.430549146285187, | |
| "grad_norm": 0.42558759450912476, | |
| "learning_rate": 1.4287911806178696e-05, | |
| "loss": 1.0759, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.4766958929395477, | |
| "grad_norm": 0.24304209649562836, | |
| "learning_rate": 1.4031534418664275e-05, | |
| "loss": 1.0637, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.5228426395939088, | |
| "grad_norm": 0.25986990332603455, | |
| "learning_rate": 1.3775157031149852e-05, | |
| "loss": 1.0446, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.5689893862482696, | |
| "grad_norm": 0.20725102722644806, | |
| "learning_rate": 1.3518779643635433e-05, | |
| "loss": 1.0384, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.6151361329026304, | |
| "grad_norm": 0.18936870992183685, | |
| "learning_rate": 1.3262402256121012e-05, | |
| "loss": 1.0139, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6612828795569912, | |
| "grad_norm": 0.19966499507427216, | |
| "learning_rate": 1.300602486860659e-05, | |
| "loss": 1.0058, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.707429626211352, | |
| "grad_norm": 0.30528759956359863, | |
| "learning_rate": 1.2749647481092169e-05, | |
| "loss": 0.9838, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.753576372865713, | |
| "grad_norm": 0.2316664308309555, | |
| "learning_rate": 1.2493270093577748e-05, | |
| "loss": 0.9904, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.7997231195200738, | |
| "grad_norm": 0.2217002511024475, | |
| "learning_rate": 1.2236892706063325e-05, | |
| "loss": 0.9735, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.8458698661744348, | |
| "grad_norm": 0.2654038369655609, | |
| "learning_rate": 1.1980515318548904e-05, | |
| "loss": 0.9804, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8920166128287956, | |
| "grad_norm": 0.20543397963047028, | |
| "learning_rate": 1.1724137931034483e-05, | |
| "loss": 0.9424, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.9381633594831564, | |
| "grad_norm": 0.24414564669132233, | |
| "learning_rate": 1.1467760543520064e-05, | |
| "loss": 0.9353, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.9843101061375172, | |
| "grad_norm": 0.19333013892173767, | |
| "learning_rate": 1.1211383156005641e-05, | |
| "loss": 0.9374, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.030456852791878, | |
| "grad_norm": 0.20063996315002441, | |
| "learning_rate": 1.095500576849122e-05, | |
| "loss": 0.9409, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.076603599446239, | |
| "grad_norm": 0.4319429397583008, | |
| "learning_rate": 1.0698628380976798e-05, | |
| "loss": 0.93, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.1227503461006, | |
| "grad_norm": 0.21358811855316162, | |
| "learning_rate": 1.0442250993462377e-05, | |
| "loss": 0.9231, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.168897092754961, | |
| "grad_norm": 0.2252470701932907, | |
| "learning_rate": 1.0185873605947956e-05, | |
| "loss": 0.9267, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.2150438394093217, | |
| "grad_norm": 0.6058911681175232, | |
| "learning_rate": 9.929496218433535e-06, | |
| "loss": 0.902, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.2611905860636825, | |
| "grad_norm": 0.27027812600135803, | |
| "learning_rate": 9.673118830919114e-06, | |
| "loss": 0.8908, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.3073373327180433, | |
| "grad_norm": 0.3116415739059448, | |
| "learning_rate": 9.416741443404692e-06, | |
| "loss": 0.8971, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.353484079372404, | |
| "grad_norm": 0.2324889898300171, | |
| "learning_rate": 9.160364055890272e-06, | |
| "loss": 0.8927, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 2.3996308260267654, | |
| "grad_norm": 0.18322697281837463, | |
| "learning_rate": 8.90398666837585e-06, | |
| "loss": 0.8844, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.445777572681126, | |
| "grad_norm": 1.4241108894348145, | |
| "learning_rate": 8.650173054736572e-06, | |
| "loss": 0.883, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 2.491924319335487, | |
| "grad_norm": 0.22246557474136353, | |
| "learning_rate": 8.393795667222153e-06, | |
| "loss": 0.8841, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.5380710659898478, | |
| "grad_norm": 0.19322210550308228, | |
| "learning_rate": 8.13741827970773e-06, | |
| "loss": 0.89, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5842178126442086, | |
| "grad_norm": 0.20664915442466736, | |
| "learning_rate": 7.881040892193309e-06, | |
| "loss": 0.8762, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.6303645592985694, | |
| "grad_norm": 0.19776581227779388, | |
| "learning_rate": 7.624663504678887e-06, | |
| "loss": 0.8649, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 2.67651130595293, | |
| "grad_norm": 0.7888526916503906, | |
| "learning_rate": 7.368286117164467e-06, | |
| "loss": 0.8678, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.722658052607291, | |
| "grad_norm": 0.7994652390480042, | |
| "learning_rate": 7.1119087296500455e-06, | |
| "loss": 0.7896, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 2.768804799261652, | |
| "grad_norm": 0.2278624325990677, | |
| "learning_rate": 6.855531342135624e-06, | |
| "loss": 0.7324, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.814951545916013, | |
| "grad_norm": 0.24736915528774261, | |
| "learning_rate": 6.599153954621203e-06, | |
| "loss": 0.7139, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 2.861098292570374, | |
| "grad_norm": 0.22141049802303314, | |
| "learning_rate": 6.342776567106782e-06, | |
| "loss": 0.7116, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.9072450392247347, | |
| "grad_norm": 0.21581608057022095, | |
| "learning_rate": 6.08639917959236e-06, | |
| "loss": 0.7185, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 2.9533917858790955, | |
| "grad_norm": 0.27274981141090393, | |
| "learning_rate": 5.830021792077939e-06, | |
| "loss": 0.7114, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.9995385325334563, | |
| "grad_norm": 0.4279099106788635, | |
| "learning_rate": 5.5736444045635175e-06, | |
| "loss": 0.6974, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.045685279187817, | |
| "grad_norm": 0.4010777473449707, | |
| "learning_rate": 5.3172670170490966e-06, | |
| "loss": 0.7089, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 3.0918320258421783, | |
| "grad_norm": 0.4470697343349457, | |
| "learning_rate": 5.060889629534676e-06, | |
| "loss": 0.6895, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 3.137978772496539, | |
| "grad_norm": 0.534737229347229, | |
| "learning_rate": 4.804512242020255e-06, | |
| "loss": 0.709, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.1841255191509, | |
| "grad_norm": 0.3858148157596588, | |
| "learning_rate": 4.548134854505833e-06, | |
| "loss": 0.7008, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 3.2302722658052607, | |
| "grad_norm": 4.884620189666748, | |
| "learning_rate": 4.291757466991412e-06, | |
| "loss": 0.6926, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.2764190124596215, | |
| "grad_norm": 0.44127726554870605, | |
| "learning_rate": 4.03538007947699e-06, | |
| "loss": 0.6986, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 3.3225657591139823, | |
| "grad_norm": 1.2448310852050781, | |
| "learning_rate": 3.7790026919625694e-06, | |
| "loss": 0.6884, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.368712505768343, | |
| "grad_norm": 0.8101204633712769, | |
| "learning_rate": 3.522625304448148e-06, | |
| "loss": 0.6851, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 3.4148592524227044, | |
| "grad_norm": 0.5153388381004333, | |
| "learning_rate": 3.2662479169337267e-06, | |
| "loss": 0.6938, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.461005999077065, | |
| "grad_norm": 0.33079493045806885, | |
| "learning_rate": 3.0098705294193053e-06, | |
| "loss": 0.6948, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.507152745731426, | |
| "grad_norm": 1.0203328132629395, | |
| "learning_rate": 2.7534931419048844e-06, | |
| "loss": 0.6855, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.553299492385787, | |
| "grad_norm": 0.3520820140838623, | |
| "learning_rate": 2.4971157543904627e-06, | |
| "loss": 0.6849, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 3.5994462390401476, | |
| "grad_norm": 0.28180956840515137, | |
| "learning_rate": 2.2407383668760417e-06, | |
| "loss": 0.699, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 3.6455929856945084, | |
| "grad_norm": 0.33973556756973267, | |
| "learning_rate": 1.9843609793616204e-06, | |
| "loss": 0.6756, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 3.6917397323488697, | |
| "grad_norm": 0.3416615128517151, | |
| "learning_rate": 1.7279835918471993e-06, | |
| "loss": 0.685, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.7378864790032305, | |
| "grad_norm": 0.7213825583457947, | |
| "learning_rate": 1.4716062043327781e-06, | |
| "loss": 0.6764, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 3.7840332256575913, | |
| "grad_norm": 0.5637441873550415, | |
| "learning_rate": 1.2152288168183566e-06, | |
| "loss": 0.6896, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 3.830179972311952, | |
| "grad_norm": 0.4190536141395569, | |
| "learning_rate": 9.588514293039355e-07, | |
| "loss": 0.675, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 3.876326718966313, | |
| "grad_norm": 0.37957823276519775, | |
| "learning_rate": 7.024740417895142e-07, | |
| "loss": 0.6767, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 3.9224734656206737, | |
| "grad_norm": 0.5404504537582397, | |
| "learning_rate": 4.4609665427509294e-07, | |
| "loss": 0.695, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.9686202122750345, | |
| "grad_norm": 14.911314964294434, | |
| "learning_rate": 1.8971926676067174e-07, | |
| "loss": 0.6794, | |
| "step": 8600 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 8668, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.884583671986995e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |