| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 100, |
| "global_step": 2703, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.001849796522382538, |
| "grad_norm": 6.96875, |
| "learning_rate": 3.690036900369004e-07, |
| "loss": 2.3841, |
| "mean_token_accuracy": 0.7142525258745637, |
| "mean_token_accuracy_utility": 0.6758671099805843, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.003699593044765076, |
| "grad_norm": 5.375, |
| "learning_rate": 7.380073800738008e-07, |
| "loss": 2.44, |
| "mean_token_accuracy": 0.7034083351239047, |
| "mean_token_accuracy_utility": 0.6787569607786812, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.005549389567147614, |
| "grad_norm": 6.375, |
| "learning_rate": 1.1070110701107011e-06, |
| "loss": 2.4799, |
| "mean_token_accuracy": 0.7192726009194199, |
| "mean_token_accuracy_utility": 0.6550342248482328, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.007399186089530152, |
| "grad_norm": 5.75, |
| "learning_rate": 1.4760147601476015e-06, |
| "loss": 2.4214, |
| "mean_token_accuracy": 0.7140314371490212, |
| "mean_token_accuracy_utility": 0.6694207940058454, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.00924898261191269, |
| "grad_norm": 6.125, |
| "learning_rate": 1.845018450184502e-06, |
| "loss": 2.4576, |
| "mean_token_accuracy": 0.6982242451551848, |
| "mean_token_accuracy_utility": 0.6743842628293004, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.011098779134295227, |
| "grad_norm": 5.8125, |
| "learning_rate": 2.2140221402214023e-06, |
| "loss": 2.4057, |
| "mean_token_accuracy": 0.6958273520424059, |
| "mean_token_accuracy_utility": 0.6815243665907542, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.012948575656677765, |
| "grad_norm": 5.0625, |
| "learning_rate": 2.5830258302583027e-06, |
| "loss": 2.3002, |
| "mean_token_accuracy": 0.7262789468861228, |
| "mean_token_accuracy_utility": 0.6783613206746433, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.014798372179060304, |
| "grad_norm": 4.5625, |
| "learning_rate": 2.952029520295203e-06, |
| "loss": 2.4301, |
| "mean_token_accuracy": 0.7190556972534548, |
| "mean_token_accuracy_utility": 0.6611446400417758, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.016648168701442843, |
| "grad_norm": 4.0, |
| "learning_rate": 3.3210332103321034e-06, |
| "loss": 2.3908, |
| "mean_token_accuracy": 0.7003236818752655, |
| "mean_token_accuracy_utility": 0.6825896271498217, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.01849796522382538, |
| "grad_norm": 2.953125, |
| "learning_rate": 3.690036900369004e-06, |
| "loss": 2.3728, |
| "mean_token_accuracy": 0.7099819567661217, |
| "mean_token_accuracy_utility": 0.6764418875658564, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.020347761746207917, |
| "grad_norm": 2.5, |
| "learning_rate": 4.059040590405905e-06, |
| "loss": 2.2322, |
| "mean_token_accuracy": 0.7276828982926953, |
| "mean_token_accuracy_utility": 0.6876368782435299, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.022197558268590455, |
| "grad_norm": 2.609375, |
| "learning_rate": 4.428044280442805e-06, |
| "loss": 2.3619, |
| "mean_token_accuracy": 0.7134193136790301, |
| "mean_token_accuracy_utility": 0.6701192762625656, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.024047354790972992, |
| "grad_norm": 2.546875, |
| "learning_rate": 4.797047970479705e-06, |
| "loss": 2.3298, |
| "mean_token_accuracy": 0.7125030479794445, |
| "mean_token_accuracy_utility": 0.6765195903085616, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.02589715131335553, |
| "grad_norm": 2.515625, |
| "learning_rate": 5.166051660516605e-06, |
| "loss": 2.2763, |
| "mean_token_accuracy": 0.7089433138153147, |
| "mean_token_accuracy_utility": 0.6932962084889147, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.02774694783573807, |
| "grad_norm": 2.21875, |
| "learning_rate": 5.535055350553506e-06, |
| "loss": 2.2677, |
| "mean_token_accuracy": 0.717442613754883, |
| "mean_token_accuracy_utility": 0.6857615011623752, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.029596744358120607, |
| "grad_norm": 2.078125, |
| "learning_rate": 5.904059040590406e-06, |
| "loss": 2.1769, |
| "mean_token_accuracy": 0.7313899609497011, |
| "mean_token_accuracy_utility": 0.6955608544137176, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.031446540880503145, |
| "grad_norm": 1.9375, |
| "learning_rate": 6.273062730627307e-06, |
| "loss": 2.2237, |
| "mean_token_accuracy": 0.7339428020095348, |
| "mean_token_accuracy_utility": 0.6802996553743116, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.033296337402885685, |
| "grad_norm": 1.984375, |
| "learning_rate": 6.642066420664207e-06, |
| "loss": 2.2217, |
| "mean_token_accuracy": 0.7236370616149195, |
| "mean_token_accuracy_utility": 0.6879919191922397, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.03514613392526822, |
| "grad_norm": 1.921875, |
| "learning_rate": 7.011070110701108e-06, |
| "loss": 2.2149, |
| "mean_token_accuracy": 0.7329576193421957, |
| "mean_token_accuracy_utility": 0.6868700788021022, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.03699593044765076, |
| "grad_norm": 2.0625, |
| "learning_rate": 7.380073800738008e-06, |
| "loss": 2.2089, |
| "mean_token_accuracy": 0.7239208938579113, |
| "mean_token_accuracy_utility": 0.6900456447898358, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.03699593044765076, |
| "eval_mean_token_accuracy": 0.7184152915826045, |
| "eval_reasoning_loss": 1.019968867301941, |
| "eval_reasoning_runtime": 29.7907, |
| "eval_reasoning_samples_per_second": 4.33, |
| "eval_reasoning_steps_per_second": 1.108, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.03699593044765076, |
| "eval_mean_token_accuracy": 0.6886413966829923, |
| "eval_utility_loss": 1.216636061668396, |
| "eval_utility_runtime": 11.085, |
| "eval_utility_samples_per_second": 4.42, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.038845726970033294, |
| "grad_norm": 2.109375, |
| "learning_rate": 7.749077490774908e-06, |
| "loss": 2.0856, |
| "mean_token_accuracy": 0.733422633442439, |
| "mean_token_accuracy_utility": 0.7052970086048445, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.040695523492415835, |
| "grad_norm": 1.8828125, |
| "learning_rate": 8.11808118081181e-06, |
| "loss": 2.1088, |
| "mean_token_accuracy": 0.7446340602952325, |
| "mean_token_accuracy_utility": 0.6888837240648127, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.042545320014798375, |
| "grad_norm": 1.9140625, |
| "learning_rate": 8.48708487084871e-06, |
| "loss": 2.1954, |
| "mean_token_accuracy": 0.737814915554515, |
| "mean_token_accuracy_utility": 0.6799374275688994, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.04439511653718091, |
| "grad_norm": 1.875, |
| "learning_rate": 8.85608856088561e-06, |
| "loss": 2.2048, |
| "mean_token_accuracy": 0.7267478866965456, |
| "mean_token_accuracy_utility": 0.6854192488848587, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.04624491305956345, |
| "grad_norm": 1.875, |
| "learning_rate": 9.22509225092251e-06, |
| "loss": 2.2247, |
| "mean_token_accuracy": 0.7212929179140846, |
| "mean_token_accuracy_utility": 0.6847587997622877, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.048094709581945984, |
| "grad_norm": 1.9296875, |
| "learning_rate": 9.59409594095941e-06, |
| "loss": 2.2003, |
| "mean_token_accuracy": 0.7314459401259678, |
| "mean_token_accuracy_utility": 0.6799583693916791, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.049944506104328525, |
| "grad_norm": 1.9296875, |
| "learning_rate": 9.963099630996312e-06, |
| "loss": 2.1247, |
| "mean_token_accuracy": 0.7349971001178317, |
| "mean_token_accuracy_utility": 0.6907869495828012, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.05179430262671106, |
| "grad_norm": 1.921875, |
| "learning_rate": 1.033210332103321e-05, |
| "loss": 2.23, |
| "mean_token_accuracy": 0.716714913987365, |
| "mean_token_accuracy_utility": 0.6875602014517539, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.0536440991490936, |
| "grad_norm": 1.9609375, |
| "learning_rate": 1.0701107011070112e-05, |
| "loss": 2.1822, |
| "mean_token_accuracy": 0.7322411232322406, |
| "mean_token_accuracy_utility": 0.6874548816833327, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.05549389567147614, |
| "grad_norm": 2.21875, |
| "learning_rate": 1.1070110701107012e-05, |
| "loss": 2.1849, |
| "mean_token_accuracy": 0.7210574172065483, |
| "mean_token_accuracy_utility": 0.6898398657385576, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.057343692193858674, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.1439114391143913e-05, |
| "loss": 2.0429, |
| "mean_token_accuracy": 0.7477014404022346, |
| "mean_token_accuracy_utility": 0.6995135342810885, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.059193488716241215, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.1808118081180812e-05, |
| "loss": 2.1815, |
| "mean_token_accuracy": 0.7320845925120626, |
| "mean_token_accuracy_utility": 0.6811719919862023, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.06104328523862375, |
| "grad_norm": 1.875, |
| "learning_rate": 1.2177121771217713e-05, |
| "loss": 2.1669, |
| "mean_token_accuracy": 0.7352879218168835, |
| "mean_token_accuracy_utility": 0.6857330080032882, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.06289308176100629, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.2546125461254614e-05, |
| "loss": 2.1123, |
| "mean_token_accuracy": 0.744195847591067, |
| "mean_token_accuracy_utility": 0.6883533895992746, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.06474287828338883, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.2915129151291515e-05, |
| "loss": 2.0851, |
| "mean_token_accuracy": 0.7365516537105917, |
| "mean_token_accuracy_utility": 0.6959731043730392, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.06659267480577137, |
| "grad_norm": 1.75, |
| "learning_rate": 1.3284132841328414e-05, |
| "loss": 2.1455, |
| "mean_token_accuracy": 0.7337763989106851, |
| "mean_token_accuracy_utility": 0.686245609005549, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.0684424713281539, |
| "grad_norm": 1.875, |
| "learning_rate": 1.3653136531365315e-05, |
| "loss": 2.0004, |
| "mean_token_accuracy": 0.7532473223667105, |
| "mean_token_accuracy_utility": 0.7006470756225327, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.07029226785053644, |
| "grad_norm": 2.0625, |
| "learning_rate": 1.4022140221402215e-05, |
| "loss": 2.0618, |
| "mean_token_accuracy": 0.7510399268479208, |
| "mean_token_accuracy_utility": 0.6913565623119153, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.07214206437291898, |
| "grad_norm": 1.921875, |
| "learning_rate": 1.4391143911439116e-05, |
| "loss": 2.146, |
| "mean_token_accuracy": 0.7344738224739542, |
| "mean_token_accuracy_utility": 0.6873232408602401, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.07399186089530152, |
| "grad_norm": 2.03125, |
| "learning_rate": 1.4760147601476015e-05, |
| "loss": 2.1003, |
| "mean_token_accuracy": 0.7388721721951235, |
| "mean_token_accuracy_utility": 0.6882846932313965, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.07399186089530152, |
| "eval_mean_token_accuracy": 0.7335134347969762, |
| "eval_reasoning_loss": 0.9337891340255737, |
| "eval_reasoning_runtime": 29.3441, |
| "eval_reasoning_samples_per_second": 4.396, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.07399186089530152, |
| "eval_mean_token_accuracy": 0.6931954824665644, |
| "eval_utility_loss": 1.1928104162216187, |
| "eval_utility_runtime": 11.0813, |
| "eval_utility_samples_per_second": 4.422, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.07584165741768406, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.5129151291512916e-05, |
| "loss": 2.1212, |
| "mean_token_accuracy": 0.7287533668194538, |
| "mean_token_accuracy_utility": 0.6938069111124248, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.07769145394006659, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.5498154981549817e-05, |
| "loss": 2.2122, |
| "mean_token_accuracy": 0.733220936863292, |
| "mean_token_accuracy_utility": 0.676245309537307, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.07954125046244913, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.5867158671586716e-05, |
| "loss": 1.9821, |
| "mean_token_accuracy": 0.7491503027843107, |
| "mean_token_accuracy_utility": 0.7040247515752383, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.08139104698483167, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.623616236162362e-05, |
| "loss": 2.1254, |
| "mean_token_accuracy": 0.7432010590042865, |
| "mean_token_accuracy_utility": 0.6832402347006175, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.08324084350721421, |
| "grad_norm": 1.8984375, |
| "learning_rate": 1.6605166051660518e-05, |
| "loss": 2.1127, |
| "mean_token_accuracy": 0.7417561522190846, |
| "mean_token_accuracy_utility": 0.6846527110697298, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.08509064002959675, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.697416974169742e-05, |
| "loss": 2.1336, |
| "mean_token_accuracy": 0.7326888955066538, |
| "mean_token_accuracy_utility": 0.6891620348191909, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.08694043655197928, |
| "grad_norm": 2.0, |
| "learning_rate": 1.734317343173432e-05, |
| "loss": 2.0822, |
| "mean_token_accuracy": 0.7518215459065247, |
| "mean_token_accuracy_utility": 0.6880156350194306, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.08879023307436182, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.771217712177122e-05, |
| "loss": 2.0901, |
| "mean_token_accuracy": 0.7388793533556003, |
| "mean_token_accuracy_utility": 0.6931366506092196, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.09064002959674436, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.8081180811808117e-05, |
| "loss": 2.0285, |
| "mean_token_accuracy": 0.743784102376655, |
| "mean_token_accuracy_utility": 0.7010689713730466, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.0924898261191269, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.845018450184502e-05, |
| "loss": 2.1476, |
| "mean_token_accuracy": 0.7307686586935654, |
| "mean_token_accuracy_utility": 0.6927701234365297, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.09433962264150944, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.8819188191881922e-05, |
| "loss": 2.0297, |
| "mean_token_accuracy": 0.7532805705739729, |
| "mean_token_accuracy_utility": 0.6943690789542806, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.09618941916389197, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.918819188191882e-05, |
| "loss": 2.0509, |
| "mean_token_accuracy": 0.7479241122096331, |
| "mean_token_accuracy_utility": 0.6938004672633061, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.09803921568627451, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.955719557195572e-05, |
| "loss": 2.1381, |
| "mean_token_accuracy": 0.7462332167452559, |
| "mean_token_accuracy_utility": 0.6814224058652727, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.09988901220865705, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.9926199261992623e-05, |
| "loss": 2.0234, |
| "mean_token_accuracy": 0.7671792391704171, |
| "mean_token_accuracy_utility": 0.6821153654916257, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.10173880873103959, |
| "grad_norm": 1.921875, |
| "learning_rate": 1.9999866506037346e-05, |
| "loss": 2.1192, |
| "mean_token_accuracy": 0.7427834637280417, |
| "mean_token_accuracy_utility": 0.6878927994590973, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.10358860525342212, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.9999324192922497e-05, |
| "loss": 2.0469, |
| "mean_token_accuracy": 0.7621012342820958, |
| "mean_token_accuracy_utility": 0.6880790269337625, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.10543840177580466, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.9998364739888954e-05, |
| "loss": 1.9936, |
| "mean_token_accuracy": 0.749417573665158, |
| "mean_token_accuracy_utility": 0.7021325298784278, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.1072881982981872, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.9996988186962044e-05, |
| "loss": 2.0011, |
| "mean_token_accuracy": 0.7610658541115638, |
| "mean_token_accuracy_utility": 0.6927568472522642, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.10913799482056974, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.999519459156716e-05, |
| "loss": 1.9709, |
| "mean_token_accuracy": 0.7559801241215183, |
| "mean_token_accuracy_utility": 0.6993481794297357, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.11098779134295228, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.9992984028527388e-05, |
| "loss": 1.9949, |
| "mean_token_accuracy": 0.7521170280873448, |
| "mean_token_accuracy_utility": 0.7005848495213534, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.11098779134295228, |
| "eval_mean_token_accuracy": 0.740788385909667, |
| "eval_reasoning_loss": 0.8936205506324768, |
| "eval_reasoning_runtime": 29.3131, |
| "eval_reasoning_samples_per_second": 4.401, |
| "eval_reasoning_steps_per_second": 1.126, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.11098779134295228, |
| "eval_mean_token_accuracy": 0.695763579055721, |
| "eval_utility_loss": 1.1802016496658325, |
| "eval_utility_runtime": 11.0667, |
| "eval_utility_samples_per_second": 4.428, |
| "eval_utility_steps_per_second": 1.175, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.1128375878653348, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.9990356590060363e-05, |
| "loss": 2.0745, |
| "mean_token_accuracy": 0.7455725365950906, |
| "mean_token_accuracy_utility": 0.6895831368585716, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.11468738438771735, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.9987312385774444e-05, |
| "loss": 1.9427, |
| "mean_token_accuracy": 0.7677331507703946, |
| "mean_token_accuracy_utility": 0.7013643706493485, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.11653718091009989, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.9983851542664125e-05, |
| "loss": 2.0154, |
| "mean_token_accuracy": 0.7548173885782149, |
| "mean_token_accuracy_utility": 0.6952112285187139, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.11838697743248243, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.9979974205104743e-05, |
| "loss": 2.0053, |
| "mean_token_accuracy": 0.7474490461217764, |
| "mean_token_accuracy_utility": 0.6978755432930793, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.12023677395486497, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.9975680534846457e-05, |
| "loss": 2.0154, |
| "mean_token_accuracy": 0.7539827450050969, |
| "mean_token_accuracy_utility": 0.6954161911342235, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.1220865704772475, |
| "grad_norm": 2.421875, |
| "learning_rate": 1.99709707110075e-05, |
| "loss": 2.0265, |
| "mean_token_accuracy": 0.7517872219165358, |
| "mean_token_accuracy_utility": 0.6908141031277719, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.12393636699963004, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.99658449300667e-05, |
| "loss": 2.0079, |
| "mean_token_accuracy": 0.7418196267262515, |
| "mean_token_accuracy_utility": 0.7030070587757853, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.12578616352201258, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.9960303405855295e-05, |
| "loss": 2.0517, |
| "mean_token_accuracy": 0.7569198547713614, |
| "mean_token_accuracy_utility": 0.6872079278054122, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.12763596004439512, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.9954346369548002e-05, |
| "loss": 1.9238, |
| "mean_token_accuracy": 0.7735624602091704, |
| "mean_token_accuracy_utility": 0.6983795450346488, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.12948575656677766, |
| "grad_norm": 1.75, |
| "learning_rate": 1.994797406965338e-05, |
| "loss": 2.0543, |
| "mean_token_accuracy": 0.740953260025613, |
| "mean_token_accuracy_utility": 0.6959896727276697, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.1313355530891602, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.9941186772003463e-05, |
| "loss": 2.0265, |
| "mean_token_accuracy": 0.7358756741277568, |
| "mean_token_accuracy_utility": 0.7034973849279192, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.13318534961154274, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.9933984759742664e-05, |
| "loss": 1.958, |
| "mean_token_accuracy": 0.7640118417925235, |
| "mean_token_accuracy_utility": 0.7002776376230997, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.13503514613392525, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.9926368333315964e-05, |
| "loss": 2.0251, |
| "mean_token_accuracy": 0.7627459178937828, |
| "mean_token_accuracy_utility": 0.6901018076480991, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.1368849426563078, |
| "grad_norm": 1.59375, |
| "learning_rate": 1.991833781045639e-05, |
| "loss": 1.9233, |
| "mean_token_accuracy": 0.766926922368178, |
| "mean_token_accuracy_utility": 0.7031338374232784, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.13873473917869034, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.9909893526171745e-05, |
| "loss": 2.0095, |
| "mean_token_accuracy": 0.7609228966419211, |
| "mean_token_accuracy_utility": 0.6933611154260604, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.14058453570107288, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.9901035832730647e-05, |
| "loss": 2.0407, |
| "mean_token_accuracy": 0.7430123252881347, |
| "mean_token_accuracy_utility": 0.7023569947272686, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.14243433222345542, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.989176509964781e-05, |
| "loss": 1.9065, |
| "mean_token_accuracy": 0.7646185311727761, |
| "mean_token_accuracy_utility": 0.7102480829634759, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.14428412874583796, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.9882081713668665e-05, |
| "loss": 2.1343, |
| "mean_token_accuracy": 0.7387666716462095, |
| "mean_token_accuracy_utility": 0.6884745389317946, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.1461339252682205, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.987198607875319e-05, |
| "loss": 2.0268, |
| "mean_token_accuracy": 0.7685933424885899, |
| "mean_token_accuracy_utility": 0.6908123907299077, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.14798372179060304, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.986147861605908e-05, |
| "loss": 1.9479, |
| "mean_token_accuracy": 0.7440390798346115, |
| "mean_token_accuracy_utility": 0.7117689727653123, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.14798372179060304, |
| "eval_mean_token_accuracy": 0.7444021380407593, |
| "eval_reasoning_loss": 0.8745293021202087, |
| "eval_reasoning_runtime": 29.3246, |
| "eval_reasoning_samples_per_second": 4.399, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.14798372179060304, |
| "eval_mean_token_accuracy": 0.6978201848969928, |
| "eval_utility_loss": 1.1730101108551025, |
| "eval_utility_runtime": 11.0761, |
| "eval_utility_samples_per_second": 4.424, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.14983351831298558, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.9850559763924176e-05, |
| "loss": 2.1002, |
| "mean_token_accuracy": 0.750752468299127, |
| "mean_token_accuracy_utility": 0.6804157205512054, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.15168331483536812, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.9839229977848164e-05, |
| "loss": 2.0575, |
| "mean_token_accuracy": 0.7417222125012298, |
| "mean_token_accuracy_utility": 0.6936137390370861, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.15353311135775063, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.9827489730473597e-05, |
| "loss": 1.9939, |
| "mean_token_accuracy": 0.7683054154559584, |
| "mean_token_accuracy_utility": 0.6935221988478472, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.15538290788013318, |
| "grad_norm": 1.65625, |
| "learning_rate": 1.9815339511566148e-05, |
| "loss": 2.0407, |
| "mean_token_accuracy": 0.7455154083485521, |
| "mean_token_accuracy_utility": 0.6916186398449595, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.15723270440251572, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.9802779827994214e-05, |
| "loss": 1.9616, |
| "mean_token_accuracy": 0.751121275760987, |
| "mean_token_accuracy_utility": 0.7054175701032996, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.15908250092489826, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.9789811203707736e-05, |
| "loss": 2.0407, |
| "mean_token_accuracy": 0.7489875265136675, |
| "mean_token_accuracy_utility": 0.692976023292025, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.1609322974472808, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.9776434179716365e-05, |
| "loss": 1.978, |
| "mean_token_accuracy": 0.7467925541905489, |
| "mean_token_accuracy_utility": 0.7033049287962233, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.16278209396966334, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.9762649314066888e-05, |
| "loss": 1.9139, |
| "mean_token_accuracy": 0.772155505473954, |
| "mean_token_accuracy_utility": 0.7057797427008209, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.16463189049204588, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.9748457181819937e-05, |
| "loss": 1.9881, |
| "mean_token_accuracy": 0.7499182744550663, |
| "mean_token_accuracy_utility": 0.7027314593266387, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.16648168701442842, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.9733858375026015e-05, |
| "loss": 2.0116, |
| "mean_token_accuracy": 0.7652818288935526, |
| "mean_token_accuracy_utility": 0.6872472935501358, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.16833148353681096, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.9718853502700783e-05, |
| "loss": 1.9972, |
| "mean_token_accuracy": 0.7536882964946757, |
| "mean_token_accuracy_utility": 0.700442389293094, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.1701812800591935, |
| "grad_norm": 1.875, |
| "learning_rate": 1.9703443190799673e-05, |
| "loss": 2.0208, |
| "mean_token_accuracy": 0.7629778966433524, |
| "mean_token_accuracy_utility": 0.6888261618929995, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.17203107658157601, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.9687628082191748e-05, |
| "loss": 2.0488, |
| "mean_token_accuracy": 0.7499959573429029, |
| "mean_token_accuracy_utility": 0.6920548265760917, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.17388087310395856, |
| "grad_norm": 1.8984375, |
| "learning_rate": 1.9671408836632915e-05, |
| "loss": 2.0175, |
| "mean_token_accuracy": 0.7705584399414922, |
| "mean_token_accuracy_utility": 0.6842695877211796, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.1757306696263411, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.9654786130738372e-05, |
| "loss": 2.0809, |
| "mean_token_accuracy": 0.7530688632888936, |
| "mean_token_accuracy_utility": 0.6865061612724963, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.17758046614872364, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.96377606579544e-05, |
| "loss": 2.0331, |
| "mean_token_accuracy": 0.7446500372142317, |
| "mean_token_accuracy_utility": 0.698887685374142, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.17943026267110618, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.9620333128529436e-05, |
| "loss": 1.9188, |
| "mean_token_accuracy": 0.7686659648010317, |
| "mean_token_accuracy_utility": 0.70187999142061, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.18128005919348872, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.9602504269484432e-05, |
| "loss": 1.9871, |
| "mean_token_accuracy": 0.7461724071523236, |
| "mean_token_accuracy_utility": 0.7078984901546421, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.18312985571587126, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.958427482458253e-05, |
| "loss": 2.0503, |
| "mean_token_accuracy": 0.7488886622983749, |
| "mean_token_accuracy_utility": 0.6902619672672261, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.1849796522382538, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.9565645554298043e-05, |
| "loss": 1.985, |
| "mean_token_accuracy": 0.7744054974779639, |
| "mean_token_accuracy_utility": 0.6895271320763341, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.1849796522382538, |
| "eval_mean_token_accuracy": 0.7467299689018266, |
| "eval_reasoning_loss": 0.8632003664970398, |
| "eval_reasoning_runtime": 29.327, |
| "eval_reasoning_samples_per_second": 4.399, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.1849796522382538, |
| "eval_mean_token_accuracy": 0.698077537913027, |
| "eval_utility_loss": 1.1687300205230713, |
| "eval_utility_runtime": 11.0805, |
| "eval_utility_samples_per_second": 4.422, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.18682944876063634, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.9546617235784716e-05, |
| "loss": 2.0112, |
| "mean_token_accuracy": 0.7431881036704896, |
| "mean_token_accuracy_utility": 0.7005303528010141, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.18867924528301888, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.952719066284333e-05, |
| "loss": 1.9762, |
| "mean_token_accuracy": 0.7556620658473965, |
| "mean_token_accuracy_utility": 0.6972599575611572, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.1905290418054014, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.9507366645888544e-05, |
| "loss": 2.0044, |
| "mean_token_accuracy": 0.7605413669944546, |
| "mean_token_accuracy_utility": 0.6906889921209186, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.19237883832778394, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.9487146011915142e-05, |
| "loss": 1.9762, |
| "mean_token_accuracy": 0.7588870552584641, |
| "mean_token_accuracy_utility": 0.6979820731613515, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.19422863485016648, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.9466529604463484e-05, |
| "loss": 2.0293, |
| "mean_token_accuracy": 0.7561408604208, |
| "mean_token_accuracy_utility": 0.6904672707113665, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.19607843137254902, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.9445518283584345e-05, |
| "loss": 1.931, |
| "mean_token_accuracy": 0.7500006645145572, |
| "mean_token_accuracy_utility": 0.71154030458266, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.19792822789493156, |
| "grad_norm": 1.6015625, |
| "learning_rate": 1.942411292580304e-05, |
| "loss": 1.9065, |
| "mean_token_accuracy": 0.7715943297349075, |
| "mean_token_accuracy_utility": 0.7025967186418038, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.1997780244173141, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.9402314424082824e-05, |
| "loss": 1.9913, |
| "mean_token_accuracy": 0.7554707748086019, |
| "mean_token_accuracy_utility": 0.6975893866663573, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.20162782093969664, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.938012368778768e-05, |
| "loss": 1.99, |
| "mean_token_accuracy": 0.7526471620173283, |
| "mean_token_accuracy_utility": 0.7024723872511551, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.20347761746207918, |
| "grad_norm": 1.75, |
| "learning_rate": 1.935754164264437e-05, |
| "loss": 1.9788, |
| "mean_token_accuracy": 0.7598793375308714, |
| "mean_token_accuracy_utility": 0.7004828981403861, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.20532741398446172, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.9334569230703794e-05, |
| "loss": 1.9539, |
| "mean_token_accuracy": 0.7580793888725631, |
| "mean_token_accuracy_utility": 0.7058805974780844, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.20717721050684423, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.9311207410301728e-05, |
| "loss": 2.0316, |
| "mean_token_accuracy": 0.7465066504329072, |
| "mean_token_accuracy_utility": 0.6977913254855733, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.20902700702922677, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.9287457156018824e-05, |
| "loss": 1.9416, |
| "mean_token_accuracy": 0.771262233843025, |
| "mean_token_accuracy_utility": 0.6979831070518216, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.21087680355160932, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.926331945863995e-05, |
| "loss": 1.9268, |
| "mean_token_accuracy": 0.7615490348942119, |
| "mean_token_accuracy_utility": 0.7074961546728148, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.21272660007399186, |
| "grad_norm": 1.75, |
| "learning_rate": 1.9238795325112867e-05, |
| "loss": 1.9678, |
| "mean_token_accuracy": 0.7614678324597344, |
| "mean_token_accuracy_utility": 0.6953139161000201, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.2145763965963744, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.9213885778506228e-05, |
| "loss": 1.8504, |
| "mean_token_accuracy": 0.764359503102718, |
| "mean_token_accuracy_utility": 0.7189560540062184, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.21642619311875694, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.9188591857966875e-05, |
| "loss": 1.9669, |
| "mean_token_accuracy": 0.756700995308621, |
| "mean_token_accuracy_utility": 0.7036004601342662, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.21827598964113948, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.916291461867652e-05, |
| "loss": 1.9006, |
| "mean_token_accuracy": 0.7639503377433965, |
| "mean_token_accuracy_utility": 0.7139992608144947, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.22012578616352202, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.9136855131807705e-05, |
| "loss": 2.0368, |
| "mean_token_accuracy": 0.7556774617096415, |
| "mean_token_accuracy_utility": 0.6896043599753023, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.22197558268590456, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.9110414484479116e-05, |
| "loss": 1.9832, |
| "mean_token_accuracy": 0.7534921287494558, |
| "mean_token_accuracy_utility": 0.701576119295144, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.22197558268590456, |
| "eval_mean_token_accuracy": 0.7481626370749476, |
| "eval_reasoning_loss": 0.8554648160934448, |
| "eval_reasoning_runtime": 29.2881, |
| "eval_reasoning_samples_per_second": 4.405, |
| "eval_reasoning_steps_per_second": 1.127, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.22197558268590456, |
| "eval_mean_token_accuracy": 0.6991555732080224, |
| "eval_utility_loss": 1.1651310920715332, |
| "eval_utility_runtime": 11.0567, |
| "eval_utility_samples_per_second": 4.432, |
| "eval_utility_steps_per_second": 1.176, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.2238253792082871, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.908359377971025e-05, |
| "loss": 2.0216, |
| "mean_token_accuracy": 0.7640225188786645, |
| "mean_token_accuracy_utility": 0.6893411740093065, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.2256751757306696, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.9056394136375377e-05, |
| "loss": 1.9774, |
| "mean_token_accuracy": 0.7659206163216518, |
| "mean_token_accuracy_utility": 0.6940803545638693, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.22752497225305215, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.902881668915688e-05, |
| "loss": 1.9455, |
| "mean_token_accuracy": 0.7551606233402386, |
| "mean_token_accuracy_utility": 0.708700241135408, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.2293747687754347, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.9000862588497918e-05, |
| "loss": 1.9139, |
| "mean_token_accuracy": 0.7620231040487303, |
| "mean_token_accuracy_utility": 0.7063941880264907, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.23122456529781724, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.897253300055443e-05, |
| "loss": 1.9265, |
| "mean_token_accuracy": 0.7580079009987992, |
| "mean_token_accuracy_utility": 0.7050462953915007, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.23307436182019978, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.8943829107146485e-05, |
| "loss": 2.0652, |
| "mean_token_accuracy": 0.752735995221744, |
| "mean_token_accuracy_utility": 0.686512731805496, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.23492415834258232, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.891475210570898e-05, |
| "loss": 1.9428, |
| "mean_token_accuracy": 0.7590939356026671, |
| "mean_token_accuracy_utility": 0.7082580379836748, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.23677395486496486, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.8885303209241694e-05, |
| "loss": 2.0013, |
| "mean_token_accuracy": 0.7511150772070805, |
| "mean_token_accuracy_utility": 0.6999827047889242, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.2386237513873474, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.8855483646258677e-05, |
| "loss": 1.9618, |
| "mean_token_accuracy": 0.7688228471250744, |
| "mean_token_accuracy_utility": 0.6896676551308012, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.24047354790972994, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.882529466073701e-05, |
| "loss": 1.9285, |
| "mean_token_accuracy": 0.7643768834778548, |
| "mean_token_accuracy_utility": 0.7005087455460225, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.24232334443211248, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.879473751206489e-05, |
| "loss": 2.0263, |
| "mean_token_accuracy": 0.7559299363274004, |
| "mean_token_accuracy_utility": 0.6899928608253901, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.244173140954495, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.876381347498913e-05, |
| "loss": 2.0008, |
| "mean_token_accuracy": 0.7558182275884706, |
| "mean_token_accuracy_utility": 0.6930457459199539, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.24602293747687753, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.8732523839561934e-05, |
| "loss": 1.9868, |
| "mean_token_accuracy": 0.7478639949344185, |
| "mean_token_accuracy_utility": 0.7028638832926714, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.24787273399926008, |
| "grad_norm": 1.9765625, |
| "learning_rate": 1.8700869911087115e-05, |
| "loss": 2.036, |
| "mean_token_accuracy": 0.7483118548170373, |
| "mean_token_accuracy_utility": 0.6965281099225976, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.24972253052164262, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.8668853010065633e-05, |
| "loss": 1.9453, |
| "mean_token_accuracy": 0.7623911710660554, |
| "mean_token_accuracy_utility": 0.699869587069941, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.25157232704402516, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.8636474472140494e-05, |
| "loss": 2.0788, |
| "mean_token_accuracy": 0.7497218625320909, |
| "mean_token_accuracy_utility": 0.684533969551831, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.2534221235664077, |
| "grad_norm": 1.875, |
| "learning_rate": 1.8603735648041054e-05, |
| "loss": 1.9546, |
| "mean_token_accuracy": 0.759317964895106, |
| "mean_token_accuracy_utility": 0.6977686760526048, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.25527192008879024, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.857063790352665e-05, |
| "loss": 1.8964, |
| "mean_token_accuracy": 0.7794884950556955, |
| "mean_token_accuracy_utility": 0.7003220539511944, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.2571217166111728, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.853718261932964e-05, |
| "loss": 1.8888, |
| "mean_token_accuracy": 0.7688903729847285, |
| "mean_token_accuracy_utility": 0.708703385302031, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.2589715131335553, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.85033711910978e-05, |
| "loss": 1.9471, |
| "mean_token_accuracy": 0.765590254907874, |
| "mean_token_accuracy_utility": 0.7034984784508899, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.2589715131335553, |
| "eval_mean_token_accuracy": 0.7491252166789526, |
| "eval_reasoning_loss": 0.8498302698135376, |
| "eval_reasoning_runtime": 29.3219, |
| "eval_reasoning_samples_per_second": 4.399, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.2589715131335553, |
| "eval_mean_token_accuracy": 0.7000045039110248, |
| "eval_utility_loss": 1.1628293991088867, |
| "eval_utility_runtime": 11.0729, |
| "eval_utility_samples_per_second": 4.425, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.26082130965593786, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.846920502933609e-05, |
| "loss": 1.9021, |
| "mean_token_accuracy": 0.7672098050635655, |
| "mean_token_accuracy_utility": 0.7062737653920031, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.2626711061783204, |
| "grad_norm": 1.75, |
| "learning_rate": 1.843468555934783e-05, |
| "loss": 1.9766, |
| "mean_token_accuracy": 0.7532319656778987, |
| "mean_token_accuracy_utility": 0.6996837580044702, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.26452090270070294, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.839981422117523e-05, |
| "loss": 1.9575, |
| "mean_token_accuracy": 0.7557241867948872, |
| "mean_token_accuracy_utility": 0.7016439052240171, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.2663706992230855, |
| "grad_norm": 1.890625, |
| "learning_rate": 1.836459246953931e-05, |
| "loss": 1.8879, |
| "mean_token_accuracy": 0.7590757913221067, |
| "mean_token_accuracy_utility": 0.7123355769089366, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.268220495745468, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.8329021773779242e-05, |
| "loss": 1.9714, |
| "mean_token_accuracy": 0.758954811056866, |
| "mean_token_accuracy_utility": 0.6997382223726225, |
| "step": 725 |
| }, |
| { |
| "epoch": 0.2700702922678505, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.8293103617791016e-05, |
| "loss": 2.022, |
| "mean_token_accuracy": 0.7552612493918166, |
| "mean_token_accuracy_utility": 0.6897869045111885, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.27192008879023305, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.825683949996556e-05, |
| "loss": 2.0229, |
| "mean_token_accuracy": 0.753056191546609, |
| "mean_token_accuracy_utility": 0.6967393745768142, |
| "step": 735 |
| }, |
| { |
| "epoch": 0.2737698853126156, |
| "grad_norm": 1.859375, |
| "learning_rate": 1.8220230933126234e-05, |
| "loss": 1.8913, |
| "mean_token_accuracy": 0.7706742931806592, |
| "mean_token_accuracy_utility": 0.705792207373848, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.27561968183499813, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.81832794444657e-05, |
| "loss": 1.9133, |
| "mean_token_accuracy": 0.7638112878612452, |
| "mean_token_accuracy_utility": 0.7094675421418029, |
| "step": 745 |
| }, |
| { |
| "epoch": 0.27746947835738067, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.814598657548223e-05, |
| "loss": 1.9266, |
| "mean_token_accuracy": 0.766421246092815, |
| "mean_token_accuracy_utility": 0.7022716086375553, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.2793192748797632, |
| "grad_norm": 2.140625, |
| "learning_rate": 1.8108353881915403e-05, |
| "loss": 1.9564, |
| "mean_token_accuracy": 0.7627688801388143, |
| "mean_token_accuracy_utility": 0.6990558377497437, |
| "step": 755 |
| }, |
| { |
| "epoch": 0.28116907140214575, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.8070382933681185e-05, |
| "loss": 1.9353, |
| "mean_token_accuracy": 0.7590690476789531, |
| "mean_token_accuracy_utility": 0.7061246028326574, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.2830188679245283, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.803207531480645e-05, |
| "loss": 1.9966, |
| "mean_token_accuracy": 0.7572792626068627, |
| "mean_token_accuracy_utility": 0.6925224721916428, |
| "step": 765 |
| }, |
| { |
| "epoch": 0.28486866444691084, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.79934326233629e-05, |
| "loss": 1.9049, |
| "mean_token_accuracy": 0.767210426621207, |
| "mean_token_accuracy_utility": 0.704673296373999, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.2867184609692934, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.7954456471400393e-05, |
| "loss": 1.9921, |
| "mean_token_accuracy": 0.7522954994181942, |
| "mean_token_accuracy_utility": 0.6953698605206351, |
| "step": 775 |
| }, |
| { |
| "epoch": 0.2885682574916759, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.7915148484879706e-05, |
| "loss": 1.9763, |
| "mean_token_accuracy": 0.7603081222361442, |
| "mean_token_accuracy_utility": 0.6976047867617131, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.29041805401405846, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.7875510303604678e-05, |
| "loss": 1.9537, |
| "mean_token_accuracy": 0.7695711062919735, |
| "mean_token_accuracy_utility": 0.7000820265152736, |
| "step": 785 |
| }, |
| { |
| "epoch": 0.292267850536441, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.783554358115384e-05, |
| "loss": 1.9498, |
| "mean_token_accuracy": 0.7462538485766637, |
| "mean_token_accuracy_utility": 0.7100101469815102, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.29411764705882354, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.7795249984811397e-05, |
| "loss": 1.9594, |
| "mean_token_accuracy": 0.7722431572640694, |
| "mean_token_accuracy_utility": 0.6940526094154679, |
| "step": 795 |
| }, |
| { |
| "epoch": 0.2959674435812061, |
| "grad_norm": 1.75, |
| "learning_rate": 1.77546311954977e-05, |
| "loss": 1.9505, |
| "mean_token_accuracy": 0.7511213124234773, |
| "mean_token_accuracy_utility": 0.705738599537488, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.2959674435812061, |
| "eval_mean_token_accuracy": 0.7502046478820066, |
| "eval_reasoning_loss": 0.8454593420028687, |
| "eval_reasoning_runtime": 29.2756, |
| "eval_reasoning_samples_per_second": 4.406, |
| "eval_reasoning_steps_per_second": 1.127, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.2959674435812061, |
| "eval_mean_token_accuracy": 0.7000852694011209, |
| "eval_utility_loss": 1.1609435081481934, |
| "eval_utility_runtime": 11.0569, |
| "eval_utility_samples_per_second": 4.432, |
| "eval_utility_steps_per_second": 1.176, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.2978172401035886, |
| "grad_norm": 1.6640625, |
| "learning_rate": 1.7713688907699107e-05, |
| "loss": 1.9903, |
| "mean_token_accuracy": 0.7437335501434683, |
| "mean_token_accuracy_utility": 0.7080256861808427, |
| "step": 805 |
| }, |
| { |
| "epoch": 0.29966703662597116, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.767242482939731e-05, |
| "loss": 1.9282, |
| "mean_token_accuracy": 0.7578269710801119, |
| "mean_token_accuracy_utility": 0.7079195829767804, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.3015168331483537, |
| "grad_norm": 2.25, |
| "learning_rate": 1.7630840681998068e-05, |
| "loss": 1.8522, |
| "mean_token_accuracy": 0.7692865728098448, |
| "mean_token_accuracy_utility": 0.7146648218174019, |
| "step": 815 |
| }, |
| { |
| "epoch": 0.30336662967073624, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.75889382002594e-05, |
| "loss": 1.9267, |
| "mean_token_accuracy": 0.7559983470385662, |
| "mean_token_accuracy_utility": 0.7083192332635321, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.3052164261931188, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.754671913221923e-05, |
| "loss": 1.9807, |
| "mean_token_accuracy": 0.7568945106850109, |
| "mean_token_accuracy_utility": 0.7016302471077509, |
| "step": 825 |
| }, |
| { |
| "epoch": 0.30706622271550127, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.7504185239122432e-05, |
| "loss": 1.8921, |
| "mean_token_accuracy": 0.7663536736397148, |
| "mean_token_accuracy_utility": 0.7071774461476249, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.3089160192378838, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.7461338295347404e-05, |
| "loss": 1.9089, |
| "mean_token_accuracy": 0.7689913290608331, |
| "mean_token_accuracy_utility": 0.7012562255188012, |
| "step": 835 |
| }, |
| { |
| "epoch": 0.31076581576026635, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.7418180088332005e-05, |
| "loss": 1.874, |
| "mean_token_accuracy": 0.7619049732448007, |
| "mean_token_accuracy_utility": 0.7154483397247094, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.3126156122826489, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.7374712418498997e-05, |
| "loss": 1.9674, |
| "mean_token_accuracy": 0.76392020778942, |
| "mean_token_accuracy_utility": 0.6982800983179358, |
| "step": 845 |
| }, |
| { |
| "epoch": 0.31446540880503143, |
| "grad_norm": 1.6640625, |
| "learning_rate": 1.7330937099180956e-05, |
| "loss": 2.0032, |
| "mean_token_accuracy": 0.7527823741887641, |
| "mean_token_accuracy_utility": 0.6971678402607633, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.316315205327414, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.7286855956544616e-05, |
| "loss": 1.9307, |
| "mean_token_accuracy": 0.7627936670521023, |
| "mean_token_accuracy_utility": 0.7057338155195245, |
| "step": 855 |
| }, |
| { |
| "epoch": 0.3181650018497965, |
| "grad_norm": 2.0, |
| "learning_rate": 1.7242470829514674e-05, |
| "loss": 1.9394, |
| "mean_token_accuracy": 0.7516862940875723, |
| "mean_token_accuracy_utility": 0.7085610944222528, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.32001479837217905, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.7197783569697084e-05, |
| "loss": 1.9544, |
| "mean_token_accuracy": 0.7622102706682473, |
| "mean_token_accuracy_utility": 0.7009024913195805, |
| "step": 865 |
| }, |
| { |
| "epoch": 0.3218645948945616, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.715279604130183e-05, |
| "loss": 1.9804, |
| "mean_token_accuracy": 0.7501057315407509, |
| "mean_token_accuracy_utility": 0.6982188839022977, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.32371439141694414, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.7107510121065138e-05, |
| "loss": 1.9881, |
| "mean_token_accuracy": 0.7643347534594208, |
| "mean_token_accuracy_utility": 0.6932728081481428, |
| "step": 875 |
| }, |
| { |
| "epoch": 0.3255641879393267, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.7061927698171187e-05, |
| "loss": 1.8708, |
| "mean_token_accuracy": 0.7785076705553833, |
| "mean_token_accuracy_utility": 0.7034102582856947, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.3274139844617092, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.7016050674173304e-05, |
| "loss": 1.9629, |
| "mean_token_accuracy": 0.7627453618086614, |
| "mean_token_accuracy_utility": 0.701070267912343, |
| "step": 885 |
| }, |
| { |
| "epoch": 0.32926378098409176, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.696988096291463e-05, |
| "loss": 2.0354, |
| "mean_token_accuracy": 0.7506153920196219, |
| "mean_token_accuracy_utility": 0.7033632172007582, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.3311135775064743, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.6923420490448298e-05, |
| "loss": 1.9353, |
| "mean_token_accuracy": 0.7593608368226583, |
| "mean_token_accuracy_utility": 0.7035052421877557, |
| "step": 895 |
| }, |
| { |
| "epoch": 0.33296337402885684, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.687667119495705e-05, |
| "loss": 1.9646, |
| "mean_token_accuracy": 0.750581744002609, |
| "mean_token_accuracy_utility": 0.7041313875245543, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.33296337402885684, |
| "eval_mean_token_accuracy": 0.7507658109499459, |
| "eval_reasoning_loss": 0.8419583439826965, |
| "eval_reasoning_runtime": 29.3314, |
| "eval_reasoning_samples_per_second": 4.398, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.33296337402885684, |
| "eval_mean_token_accuracy": 0.7006531375436355, |
| "eval_utility_loss": 1.1593377590179443, |
| "eval_utility_runtime": 11.081, |
| "eval_utility_samples_per_second": 4.422, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.3348131705512394, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.6829635026672432e-05, |
| "loss": 1.8893, |
| "mean_token_accuracy": 0.7634673856790564, |
| "mean_token_accuracy_utility": 0.7105651879067187, |
| "step": 905 |
| }, |
| { |
| "epoch": 0.3366629670736219, |
| "grad_norm": 2.15625, |
| "learning_rate": 1.6782313947793377e-05, |
| "loss": 1.9199, |
| "mean_token_accuracy": 0.7593443175200326, |
| "mean_token_accuracy_utility": 0.7114361746231885, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.33851276359600446, |
| "grad_norm": 1.75, |
| "learning_rate": 1.6734709932404404e-05, |
| "loss": 1.9503, |
| "mean_token_accuracy": 0.7594987338622843, |
| "mean_token_accuracy_utility": 0.6993310414279501, |
| "step": 915 |
| }, |
| { |
| "epoch": 0.340362560118387, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.668682496639323e-05, |
| "loss": 2.0123, |
| "mean_token_accuracy": 0.7584040038839535, |
| "mean_token_accuracy_utility": 0.6944285717501011, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.3422123566407695, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.663866104736793e-05, |
| "loss": 1.922, |
| "mean_token_accuracy": 0.7749290877302911, |
| "mean_token_accuracy_utility": 0.6991778164525412, |
| "step": 925 |
| }, |
| { |
| "epoch": 0.34406215316315203, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.6590220184573623e-05, |
| "loss": 2.043, |
| "mean_token_accuracy": 0.746720949890744, |
| "mean_token_accuracy_utility": 0.6927258432157208, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.34591194968553457, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.6541504398808633e-05, |
| "loss": 1.9672, |
| "mean_token_accuracy": 0.7600060216024745, |
| "mean_token_accuracy_utility": 0.6983739656011053, |
| "step": 935 |
| }, |
| { |
| "epoch": 0.3477617462079171, |
| "grad_norm": 1.625, |
| "learning_rate": 1.6492515722340193e-05, |
| "loss": 1.9414, |
| "mean_token_accuracy": 0.7437350358410825, |
| "mean_token_accuracy_utility": 0.7143913921564267, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.34961154273029965, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.6443256198819665e-05, |
| "loss": 2.0026, |
| "mean_token_accuracy": 0.7501797872609364, |
| "mean_token_accuracy_utility": 0.698634131734871, |
| "step": 945 |
| }, |
| { |
| "epoch": 0.3514613392526822, |
| "grad_norm": 1.6640625, |
| "learning_rate": 1.6393727883197292e-05, |
| "loss": 1.9492, |
| "mean_token_accuracy": 0.7609356457400656, |
| "mean_token_accuracy_utility": 0.70094787392198, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.35331113577506473, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.6343932841636455e-05, |
| "loss": 1.8671, |
| "mean_token_accuracy": 0.756381664944222, |
| "mean_token_accuracy_utility": 0.7268877914383569, |
| "step": 955 |
| }, |
| { |
| "epoch": 0.3551609322974473, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.6293873151427505e-05, |
| "loss": 1.9966, |
| "mean_token_accuracy": 0.7538782521827635, |
| "mean_token_accuracy_utility": 0.6962052918300946, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.3570107288198298, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.6243550900901076e-05, |
| "loss": 2.0142, |
| "mean_token_accuracy": 0.7505413450954848, |
| "mean_token_accuracy_utility": 0.700890398042578, |
| "step": 965 |
| }, |
| { |
| "epoch": 0.35886052534221236, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.6192968189340993e-05, |
| "loss": 1.962, |
| "mean_token_accuracy": 0.7564054078985343, |
| "mean_token_accuracy_utility": 0.6992705421130874, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.3607103218645949, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.6142127126896682e-05, |
| "loss": 1.9821, |
| "mean_token_accuracy": 0.7505374707735399, |
| "mean_token_accuracy_utility": 0.706866429601077, |
| "step": 975 |
| }, |
| { |
| "epoch": 0.36256011838697744, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.6091029834495133e-05, |
| "loss": 2.0098, |
| "mean_token_accuracy": 0.765791667409156, |
| "mean_token_accuracy_utility": 0.6925037690859814, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.36440991490936, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.603967844375245e-05, |
| "loss": 1.9821, |
| "mean_token_accuracy": 0.7583829451026997, |
| "mean_token_accuracy_utility": 0.6974991964223969, |
| "step": 985 |
| }, |
| { |
| "epoch": 0.3662597114317425, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.59880750968849e-05, |
| "loss": 1.9105, |
| "mean_token_accuracy": 0.7601516745676019, |
| "mean_token_accuracy_utility": 0.7092186611635392, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.36810950795412506, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.593622194661956e-05, |
| "loss": 1.9195, |
| "mean_token_accuracy": 0.7741152536943507, |
| "mean_token_accuracy_utility": 0.6987505496233509, |
| "step": 995 |
| }, |
| { |
| "epoch": 0.3699593044765076, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.5884121156104496e-05, |
| "loss": 1.9324, |
| "mean_token_accuracy": 0.7608190174871624, |
| "mean_token_accuracy_utility": 0.7048910292809317, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.3699593044765076, |
| "eval_mean_token_accuracy": 0.7516358650351678, |
| "eval_reasoning_loss": 0.8389561176300049, |
| "eval_reasoning_runtime": 29.3148, |
| "eval_reasoning_samples_per_second": 4.401, |
| "eval_reasoning_steps_per_second": 1.126, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.3699593044765076, |
| "eval_mean_token_accuracy": 0.700894104816424, |
| "eval_utility_loss": 1.1582527160644531, |
| "eval_utility_runtime": 11.0763, |
| "eval_utility_samples_per_second": 4.424, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.37180910099889014, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.5831774898818558e-05, |
| "loss": 2.0622, |
| "mean_token_accuracy": 0.750454999291437, |
| "mean_token_accuracy_utility": 0.6876950946406394, |
| "step": 1005 |
| }, |
| { |
| "epoch": 0.3736588975212727, |
| "grad_norm": 1.9609375, |
| "learning_rate": 1.5779185358480666e-05, |
| "loss": 1.9597, |
| "mean_token_accuracy": 0.7589121822109223, |
| "mean_token_accuracy_utility": 0.7032747445094583, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.3755086940436552, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.5726354728958736e-05, |
| "loss": 1.9467, |
| "mean_token_accuracy": 0.7639028862797762, |
| "mean_token_accuracy_utility": 0.7034423487591641, |
| "step": 1015 |
| }, |
| { |
| "epoch": 0.37735849056603776, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.5673285214178176e-05, |
| "loss": 1.9925, |
| "mean_token_accuracy": 0.7605182489922861, |
| "mean_token_accuracy_utility": 0.6934493219824234, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.37920828708842025, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.5619979028029898e-05, |
| "loss": 1.9061, |
| "mean_token_accuracy": 0.7646967944561304, |
| "mean_token_accuracy_utility": 0.7082038480958938, |
| "step": 1025 |
| }, |
| { |
| "epoch": 0.3810580836108028, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.5566438394278005e-05, |
| "loss": 1.9864, |
| "mean_token_accuracy": 0.7726683511742358, |
| "mean_token_accuracy_utility": 0.6840123057298346, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.38290788013318533, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.5512665546467008e-05, |
| "loss": 2.1569, |
| "mean_token_accuracy": 0.7426976255859522, |
| "mean_token_accuracy_utility": 0.6837188974890722, |
| "step": 1035 |
| }, |
| { |
| "epoch": 0.38475767665556787, |
| "grad_norm": 1.8828125, |
| "learning_rate": 1.5458662727828633e-05, |
| "loss": 2.0146, |
| "mean_token_accuracy": 0.7641871639052576, |
| "mean_token_accuracy_utility": 0.6905433665292264, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.3866074731779504, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.540443219118827e-05, |
| "loss": 1.936, |
| "mean_token_accuracy": 0.7568844523731755, |
| "mean_token_accuracy_utility": 0.7049744165774279, |
| "step": 1045 |
| }, |
| { |
| "epoch": 0.38845726970033295, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.5349976198870974e-05, |
| "loss": 1.9291, |
| "mean_token_accuracy": 0.758489840407233, |
| "mean_token_accuracy_utility": 0.7108876157508812, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.3903070662227155, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.529529702260709e-05, |
| "loss": 2.0074, |
| "mean_token_accuracy": 0.756161966809062, |
| "mean_token_accuracy_utility": 0.6909742106170594, |
| "step": 1055 |
| }, |
| { |
| "epoch": 0.39215686274509803, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.5240396943437483e-05, |
| "loss": 1.9679, |
| "mean_token_accuracy": 0.751178918329906, |
| "mean_token_accuracy_utility": 0.7073703981682207, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.3940066592674806, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.5185278251618391e-05, |
| "loss": 1.9529, |
| "mean_token_accuracy": 0.756660935438817, |
| "mean_token_accuracy_utility": 0.7052392269912451, |
| "step": 1065 |
| }, |
| { |
| "epoch": 0.3958564557898631, |
| "grad_norm": 1.75, |
| "learning_rate": 1.5129943246525871e-05, |
| "loss": 1.9766, |
| "mean_token_accuracy": 0.7538348353469612, |
| "mean_token_accuracy_utility": 0.7009662048111657, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.39770625231224566, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.5074394236559871e-05, |
| "loss": 1.9182, |
| "mean_token_accuracy": 0.775512270389275, |
| "mean_token_accuracy_utility": 0.6968214257864719, |
| "step": 1075 |
| }, |
| { |
| "epoch": 0.3995560488346282, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.5018633539047955e-05, |
| "loss": 2.1118, |
| "mean_token_accuracy": 0.7535106302031561, |
| "mean_token_accuracy_utility": 0.6861145202686507, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.40140584535701074, |
| "grad_norm": 1.625, |
| "learning_rate": 1.4962663480148606e-05, |
| "loss": 1.9231, |
| "mean_token_accuracy": 0.7693060720657021, |
| "mean_token_accuracy_utility": 0.701963417326039, |
| "step": 1085 |
| }, |
| { |
| "epoch": 0.4032556418793933, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.4906486394754198e-05, |
| "loss": 1.9467, |
| "mean_token_accuracy": 0.7697517443900798, |
| "mean_token_accuracy_utility": 0.6957109574249134, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.4051054384017758, |
| "grad_norm": 1.90625, |
| "learning_rate": 1.4850104626393598e-05, |
| "loss": 1.9584, |
| "mean_token_accuracy": 0.7687631585033768, |
| "mean_token_accuracy_utility": 0.6945691159479163, |
| "step": 1095 |
| }, |
| { |
| "epoch": 0.40695523492415836, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.4793520527134381e-05, |
| "loss": 1.9476, |
| "mean_token_accuracy": 0.7659165684082223, |
| "mean_token_accuracy_utility": 0.7012963674655746, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.40695523492415836, |
| "eval_mean_token_accuracy": 0.7519687117082632, |
| "eval_reasoning_loss": 0.8368719816207886, |
| "eval_reasoning_runtime": 29.3908, |
| "eval_reasoning_samples_per_second": 4.389, |
| "eval_reasoning_steps_per_second": 1.123, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.40695523492415836, |
| "eval_mean_token_accuracy": 0.700876013551517, |
| "eval_utility_loss": 1.1573108434677124, |
| "eval_utility_runtime": 11.0998, |
| "eval_utility_samples_per_second": 4.414, |
| "eval_utility_steps_per_second": 1.171, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.4088050314465409, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.473673645748473e-05, |
| "loss": 1.9296, |
| "mean_token_accuracy": 0.7598124546818168, |
| "mean_token_accuracy_utility": 0.707799129897242, |
| "step": 1105 |
| }, |
| { |
| "epoch": 0.41065482796892344, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.4679754786294954e-05, |
| "loss": 1.9739, |
| "mean_token_accuracy": 0.7636582447580822, |
| "mean_token_accuracy_utility": 0.698118357175088, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.412504624491306, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.4622577890658668e-05, |
| "loss": 1.9209, |
| "mean_token_accuracy": 0.7663410265937536, |
| "mean_token_accuracy_utility": 0.703174416070772, |
| "step": 1115 |
| }, |
| { |
| "epoch": 0.41435442101368847, |
| "grad_norm": 1.8515625, |
| "learning_rate": 1.4565208155813626e-05, |
| "loss": 1.9249, |
| "mean_token_accuracy": 0.7656286472985618, |
| "mean_token_accuracy_utility": 0.7065398682293476, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.416204217536071, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.4507647975042221e-05, |
| "loss": 1.937, |
| "mean_token_accuracy": 0.7595634027747691, |
| "mean_token_accuracy_utility": 0.7046512277918906, |
| "step": 1125 |
| }, |
| { |
| "epoch": 0.41805401405845355, |
| "grad_norm": 1.75, |
| "learning_rate": 1.4449899749571643e-05, |
| "loss": 1.8077, |
| "mean_token_accuracy": 0.7627601367250157, |
| "mean_token_accuracy_utility": 0.7359524978823371, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.4199038105808361, |
| "grad_norm": 1.65625, |
| "learning_rate": 1.4391965888473705e-05, |
| "loss": 1.8742, |
| "mean_token_accuracy": 0.7673650965676388, |
| "mean_token_accuracy_utility": 0.7130990036784977, |
| "step": 1135 |
| }, |
| { |
| "epoch": 0.42175360710321863, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.4333848808564353e-05, |
| "loss": 1.8598, |
| "mean_token_accuracy": 0.7591250508255545, |
| "mean_token_accuracy_utility": 0.7170590005035987, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.42360340362560117, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.4275550934302822e-05, |
| "loss": 1.9453, |
| "mean_token_accuracy": 0.7702944366128612, |
| "mean_token_accuracy_utility": 0.6967502164594692, |
| "step": 1145 |
| }, |
| { |
| "epoch": 0.4254532001479837, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.4217074697690532e-05, |
| "loss": 2.0559, |
| "mean_token_accuracy": 0.7544849870048204, |
| "mean_token_accuracy_utility": 0.6855856459039502, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.42730299667036625, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.4158422538169596e-05, |
| "loss": 1.9122, |
| "mean_token_accuracy": 0.7563592058994779, |
| "mean_token_accuracy_utility": 0.7132535682802491, |
| "step": 1155 |
| }, |
| { |
| "epoch": 0.4291527931927488, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.4099596902521073e-05, |
| "loss": 1.9459, |
| "mean_token_accuracy": 0.7584998801325895, |
| "mean_token_accuracy_utility": 0.6990279511570083, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.43100258971513133, |
| "grad_norm": 1.6640625, |
| "learning_rate": 1.40406002447629e-05, |
| "loss": 1.8955, |
| "mean_token_accuracy": 0.7523480838091323, |
| "mean_token_accuracy_utility": 0.7126196948198623, |
| "step": 1165 |
| }, |
| { |
| "epoch": 0.4328523862375139, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.39814350260475e-05, |
| "loss": 1.9175, |
| "mean_token_accuracy": 0.7631545753730471, |
| "mean_token_accuracy_utility": 0.7029532950461622, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.4347021827598964, |
| "grad_norm": 2.0, |
| "learning_rate": 1.392210371455913e-05, |
| "loss": 1.9157, |
| "mean_token_accuracy": 0.7599155800197056, |
| "mean_token_accuracy_utility": 0.7089052968786257, |
| "step": 1175 |
| }, |
| { |
| "epoch": 0.43655197928227896, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.3862608785410903e-05, |
| "loss": 1.9803, |
| "mean_token_accuracy": 0.7663748973943884, |
| "mean_token_accuracy_utility": 0.6920054950982586, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.4384017758046615, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.3802952720541543e-05, |
| "loss": 1.8303, |
| "mean_token_accuracy": 0.7758111766111446, |
| "mean_token_accuracy_utility": 0.716692742068022, |
| "step": 1185 |
| }, |
| { |
| "epoch": 0.44025157232704404, |
| "grad_norm": 1.65625, |
| "learning_rate": 1.3743138008611838e-05, |
| "loss": 1.8991, |
| "mean_token_accuracy": 0.7640231565341371, |
| "mean_token_accuracy_utility": 0.7080635102171223, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.4421013688494266, |
| "grad_norm": 1.6171875, |
| "learning_rate": 1.3683167144900833e-05, |
| "loss": 1.9221, |
| "mean_token_accuracy": 0.7661825196328257, |
| "mean_token_accuracy_utility": 0.7014547807491858, |
| "step": 1195 |
| }, |
| { |
| "epoch": 0.4439511653718091, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.3623042631201713e-05, |
| "loss": 1.9099, |
| "mean_token_accuracy": 0.7498149804662443, |
| "mean_token_accuracy_utility": 0.7161688462802707, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.4439511653718091, |
| "eval_mean_token_accuracy": 0.7522725445187949, |
| "eval_reasoning_loss": 0.8353175520896912, |
| "eval_reasoning_runtime": 29.362, |
| "eval_reasoning_samples_per_second": 4.393, |
| "eval_reasoning_steps_per_second": 1.124, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.4439511653718091, |
| "eval_mean_token_accuracy": 0.7008033960006647, |
| "eval_utility_loss": 1.1565465927124023, |
| "eval_utility_runtime": 11.1051, |
| "eval_utility_samples_per_second": 4.412, |
| "eval_utility_steps_per_second": 1.171, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.44580096189419166, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.3562766975717468e-05, |
| "loss": 1.9261, |
| "mean_token_accuracy": 0.7602286826184391, |
| "mean_token_accuracy_utility": 0.7044282663001915, |
| "step": 1205 |
| }, |
| { |
| "epoch": 0.4476507584165742, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.3502342692956229e-05, |
| "loss": 1.8244, |
| "mean_token_accuracy": 0.7834447286474139, |
| "mean_token_accuracy_utility": 0.7087490354569055, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.44950055493895674, |
| "grad_norm": 1.796875, |
| "learning_rate": 1.3441772303626387e-05, |
| "loss": 1.9662, |
| "mean_token_accuracy": 0.753798327163304, |
| "mean_token_accuracy_utility": 0.7012089876723968, |
| "step": 1215 |
| }, |
| { |
| "epoch": 0.4513503514613392, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.3381058334531436e-05, |
| "loss": 1.8524, |
| "mean_token_accuracy": 0.7778755058732604, |
| "mean_token_accuracy_utility": 0.7054467214809371, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.45320014798372177, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.3320203318464552e-05, |
| "loss": 1.921, |
| "mean_token_accuracy": 0.7585274606647261, |
| "mean_token_accuracy_utility": 0.7087699013025723, |
| "step": 1225 |
| }, |
| { |
| "epoch": 0.4550499445061043, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.3259209794102952e-05, |
| "loss": 1.8348, |
| "mean_token_accuracy": 0.7642608352474359, |
| "mean_token_accuracy_utility": 0.723610374224015, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.45689974102848685, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.319808030590197e-05, |
| "loss": 1.9926, |
| "mean_token_accuracy": 0.7661896047374416, |
| "mean_token_accuracy_utility": 0.6925039249961772, |
| "step": 1235 |
| }, |
| { |
| "epoch": 0.4587495375508694, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.3136817403988918e-05, |
| "loss": 1.8545, |
| "mean_token_accuracy": 0.756519208990281, |
| "mean_token_accuracy_utility": 0.7234768871910376, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.46059933407325193, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.3075423644056699e-05, |
| "loss": 2.0352, |
| "mean_token_accuracy": 0.7442233505831997, |
| "mean_token_accuracy_utility": 0.6953549699475948, |
| "step": 1245 |
| }, |
| { |
| "epoch": 0.46244913059563447, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.3013901587257207e-05, |
| "loss": 1.9056, |
| "mean_token_accuracy": 0.758973462235769, |
| "mean_token_accuracy_utility": 0.7118688904694503, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.464298927118017, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.2952253800094467e-05, |
| "loss": 1.968, |
| "mean_token_accuracy": 0.7593009873474792, |
| "mean_token_accuracy_utility": 0.6996989377808366, |
| "step": 1255 |
| }, |
| { |
| "epoch": 0.46614872364039955, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.2890482854317565e-05, |
| "loss": 2.0867, |
| "mean_token_accuracy": 0.7403377675643255, |
| "mean_token_accuracy_utility": 0.6925068702328316, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.4679985201627821, |
| "grad_norm": 1.625, |
| "learning_rate": 1.2828591326813382e-05, |
| "loss": 1.9336, |
| "mean_token_accuracy": 0.7538210106654373, |
| "mean_token_accuracy_utility": 0.7104893584518952, |
| "step": 1265 |
| }, |
| { |
| "epoch": 0.46984831668516464, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.2766581799499078e-05, |
| "loss": 1.936, |
| "mean_token_accuracy": 0.761194664526556, |
| "mean_token_accuracy_utility": 0.707546940649115, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.4716981132075472, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.2704456859214397e-05, |
| "loss": 2.0882, |
| "mean_token_accuracy": 0.747681228709687, |
| "mean_token_accuracy_utility": 0.685594718180121, |
| "step": 1275 |
| }, |
| { |
| "epoch": 0.4735479097299297, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.264221909761374e-05, |
| "loss": 1.892, |
| "mean_token_accuracy": 0.7801635580493088, |
| "mean_token_accuracy_utility": 0.699062289543998, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.47539770625231226, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.2579871111058042e-05, |
| "loss": 1.9494, |
| "mean_token_accuracy": 0.7507679963876546, |
| "mean_token_accuracy_utility": 0.7084282001387543, |
| "step": 1285 |
| }, |
| { |
| "epoch": 0.4772475027746948, |
| "grad_norm": 1.6484375, |
| "learning_rate": 1.251741550050649e-05, |
| "loss": 1.8262, |
| "mean_token_accuracy": 0.769967734355047, |
| "mean_token_accuracy_utility": 0.7242723621659615, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.47909729929707734, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.2454854871407993e-05, |
| "loss": 1.9738, |
| "mean_token_accuracy": 0.7561046252198013, |
| "mean_token_accuracy_utility": 0.7044322090241071, |
| "step": 1295 |
| }, |
| { |
| "epoch": 0.4809470958194599, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.2392191833592495e-05, |
| "loss": 1.9205, |
| "mean_token_accuracy": 0.7689955637372224, |
| "mean_token_accuracy_utility": 0.7021949098834636, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.4809470958194599, |
| "eval_mean_token_accuracy": 0.7527017060087405, |
| "eval_reasoning_loss": 0.8336358070373535, |
| "eval_reasoning_runtime": 29.3447, |
| "eval_reasoning_samples_per_second": 4.396, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.4809470958194599, |
| "eval_mean_token_accuracy": 0.700989106331151, |
| "eval_utility_loss": 1.1558451652526855, |
| "eval_utility_runtime": 11.0967, |
| "eval_utility_samples_per_second": 4.416, |
| "eval_utility_steps_per_second": 1.172, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.4827968923418424, |
| "grad_norm": 1.75, |
| "learning_rate": 1.2329429001162114e-05, |
| "loss": 1.9014, |
| "mean_token_accuracy": 0.7665251333015027, |
| "mean_token_accuracy_utility": 0.7049273904301605, |
| "step": 1305 |
| }, |
| { |
| "epoch": 0.48464668886422496, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.226656899238207e-05, |
| "loss": 1.8372, |
| "mean_token_accuracy": 0.7723029599175455, |
| "mean_token_accuracy_utility": 0.7137894000418903, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.48649648538660745, |
| "grad_norm": 1.875, |
| "learning_rate": 1.2203614429571475e-05, |
| "loss": 1.9188, |
| "mean_token_accuracy": 0.7508753341003931, |
| "mean_token_accuracy_utility": 0.7173081405281552, |
| "step": 1315 |
| }, |
| { |
| "epoch": 0.48834628190899, |
| "grad_norm": 1.75, |
| "learning_rate": 1.214056793899394e-05, |
| "loss": 1.8335, |
| "mean_token_accuracy": 0.7640378212769015, |
| "mean_token_accuracy_utility": 0.7179710927780121, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.49019607843137253, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.207743215074801e-05, |
| "loss": 1.9036, |
| "mean_token_accuracy": 0.7686437309670915, |
| "mean_token_accuracy_utility": 0.7043587014669574, |
| "step": 1325 |
| }, |
| { |
| "epoch": 0.49204587495375507, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.2014209698657424e-05, |
| "loss": 1.8741, |
| "mean_token_accuracy": 0.7577807085519066, |
| "mean_token_accuracy_utility": 0.7167161218309537, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.4938956714761376, |
| "grad_norm": 1.9296875, |
| "learning_rate": 1.1950903220161286e-05, |
| "loss": 1.9886, |
| "mean_token_accuracy": 0.7559681095726978, |
| "mean_token_accuracy_utility": 0.6937933147977443, |
| "step": 1335 |
| }, |
| { |
| "epoch": 0.49574546799852015, |
| "grad_norm": 1.65625, |
| "learning_rate": 1.1887515356203994e-05, |
| "loss": 1.9883, |
| "mean_token_accuracy": 0.7555886327270003, |
| "mean_token_accuracy_utility": 0.6996168487642439, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.4975952645209027, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.1824048751125101e-05, |
| "loss": 1.9349, |
| "mean_token_accuracy": 0.7545376545391549, |
| "mean_token_accuracy_utility": 0.7052069718660329, |
| "step": 1345 |
| }, |
| { |
| "epoch": 0.49944506104328523, |
| "grad_norm": 1.7890625, |
| "learning_rate": 1.1760506052548984e-05, |
| "loss": 1.95, |
| "mean_token_accuracy": 0.7549141535127344, |
| "mean_token_accuracy_utility": 0.7094437625450813, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.5012948575656678, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.1696889911274394e-05, |
| "loss": 1.9173, |
| "mean_token_accuracy": 0.7651928594862293, |
| "mean_token_accuracy_utility": 0.7033898934969113, |
| "step": 1355 |
| }, |
| { |
| "epoch": 0.5031446540880503, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.1633202981163882e-05, |
| "loss": 1.9546, |
| "mean_token_accuracy": 0.7544123288610229, |
| "mean_token_accuracy_utility": 0.7093412079633425, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.5049944506104328, |
| "grad_norm": 1.6015625, |
| "learning_rate": 1.1569447919033086e-05, |
| "loss": 1.8634, |
| "mean_token_accuracy": 0.7542238964180016, |
| "mean_token_accuracy_utility": 0.7206284013589666, |
| "step": 1365 |
| }, |
| { |
| "epoch": 0.5068442471328154, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.1505627384539888e-05, |
| "loss": 1.9897, |
| "mean_token_accuracy": 0.7607402502986541, |
| "mean_token_accuracy_utility": 0.6918724347376755, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.5086940436551979, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.1441744040073469e-05, |
| "loss": 1.9375, |
| "mean_token_accuracy": 0.7539845810685265, |
| "mean_token_accuracy_utility": 0.7054206306496165, |
| "step": 1375 |
| }, |
| { |
| "epoch": 0.5105438401775805, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.1377800550643246e-05, |
| "loss": 1.9783, |
| "mean_token_accuracy": 0.7612446999474554, |
| "mean_token_accuracy_utility": 0.6978920038029467, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.512393636699963, |
| "grad_norm": 1.765625, |
| "learning_rate": 1.1313799583767693e-05, |
| "loss": 1.8298, |
| "mean_token_accuracy": 0.7631493267528319, |
| "mean_token_accuracy_utility": 0.7206457154537527, |
| "step": 1385 |
| }, |
| { |
| "epoch": 0.5142434332223456, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.124974380936305e-05, |
| "loss": 1.9374, |
| "mean_token_accuracy": 0.7572365400196988, |
| "mean_token_accuracy_utility": 0.7026835122888202, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.516093229744728, |
| "grad_norm": 1.828125, |
| "learning_rate": 1.1185635899631963e-05, |
| "loss": 1.9896, |
| "mean_token_accuracy": 0.7589725612851247, |
| "mean_token_accuracy_utility": 0.7009209610647031, |
| "step": 1395 |
| }, |
| { |
| "epoch": 0.5179430262671106, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.1121478528951994e-05, |
| "loss": 1.8746, |
| "mean_token_accuracy": 0.7658286561129081, |
| "mean_token_accuracy_utility": 0.7165308923884524, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.5179430262671106, |
| "eval_mean_token_accuracy": 0.7526459496557207, |
| "eval_reasoning_loss": 0.8326225280761719, |
| "eval_reasoning_runtime": 29.3384, |
| "eval_reasoning_samples_per_second": 4.397, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.5179430262671106, |
| "eval_mean_token_accuracy": 0.701475443444521, |
| "eval_utility_loss": 1.1555927991867065, |
| "eval_utility_runtime": 11.081, |
| "eval_utility_samples_per_second": 4.422, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.5197928227894931, |
| "grad_norm": 1.6328125, |
| "learning_rate": 1.1057274373764056e-05, |
| "loss": 1.8825, |
| "mean_token_accuracy": 0.7632881051767025, |
| "mean_token_accuracy_utility": 0.7129833725178004, |
| "step": 1405 |
| }, |
| { |
| "epoch": 0.5216426193118757, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.0993026112460765e-05, |
| "loss": 1.9204, |
| "mean_token_accuracy": 0.7612241971188808, |
| "mean_token_accuracy_utility": 0.7036907093033327, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.5234924158342582, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.0928736425274702e-05, |
| "loss": 1.9194, |
| "mean_token_accuracy": 0.7608630112603146, |
| "mean_token_accuracy_utility": 0.7046462000283218, |
| "step": 1415 |
| }, |
| { |
| "epoch": 0.5253422123566408, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.0864407994166609e-05, |
| "loss": 1.9495, |
| "mean_token_accuracy": 0.7616248592384479, |
| "mean_token_accuracy_utility": 0.7040152136619449, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.5271920088790233, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.0800043502713486e-05, |
| "loss": 1.8809, |
| "mean_token_accuracy": 0.7611502029989615, |
| "mean_token_accuracy_utility": 0.711091979024852, |
| "step": 1425 |
| }, |
| { |
| "epoch": 0.5290418054014059, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.0735645635996676e-05, |
| "loss": 2.0101, |
| "mean_token_accuracy": 0.7554223735683625, |
| "mean_token_accuracy_utility": 0.6895342060553066, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.5308916019237884, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.0671217080489816e-05, |
| "loss": 1.8859, |
| "mean_token_accuracy": 0.7671440529143382, |
| "mean_token_accuracy_utility": 0.708099065352564, |
| "step": 1435 |
| }, |
| { |
| "epoch": 0.532741398446171, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.0606760523946789e-05, |
| "loss": 1.8641, |
| "mean_token_accuracy": 0.7771187125316381, |
| "mean_token_accuracy_utility": 0.704528164391733, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.5345911949685535, |
| "grad_norm": 1.6171875, |
| "learning_rate": 1.0542278655289588e-05, |
| "loss": 1.9149, |
| "mean_token_accuracy": 0.7635880720884839, |
| "mean_token_accuracy_utility": 0.7025934922874051, |
| "step": 1445 |
| }, |
| { |
| "epoch": 0.536440991490936, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.047777416449615e-05, |
| "loss": 1.979, |
| "mean_token_accuracy": 0.7646409219697723, |
| "mean_token_accuracy_utility": 0.6958453764805942, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.5382907880133185, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.0413249742488132e-05, |
| "loss": 1.9364, |
| "mean_token_accuracy": 0.7621164266018728, |
| "mean_token_accuracy_utility": 0.7007962263882046, |
| "step": 1455 |
| }, |
| { |
| "epoch": 0.540140584535701, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.034870808101866e-05, |
| "loss": 1.8714, |
| "mean_token_accuracy": 0.7634056950358439, |
| "mean_token_accuracy_utility": 0.7136155642795045, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.5419903810580836, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.0284151872560042e-05, |
| "loss": 2.0095, |
| "mean_token_accuracy": 0.760958340718338, |
| "mean_token_accuracy_utility": 0.6974862634406465, |
| "step": 1465 |
| }, |
| { |
| "epoch": 0.5438401775804661, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.0219583810191422e-05, |
| "loss": 1.925, |
| "mean_token_accuracy": 0.7661323746837362, |
| "mean_token_accuracy_utility": 0.7040352124254204, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.5456899741028487, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.0155006587486468e-05, |
| "loss": 1.9897, |
| "mean_token_accuracy": 0.7489752837458287, |
| "mean_token_accuracy_utility": 0.7013317942415316, |
| "step": 1475 |
| }, |
| { |
| "epoch": 0.5475397706252312, |
| "grad_norm": 1.7421875, |
| "learning_rate": 1.0090422898400985e-05, |
| "loss": 2.0054, |
| "mean_token_accuracy": 0.7640254228546554, |
| "mean_token_accuracy_utility": 0.6906453829043702, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.5493895671476138, |
| "grad_norm": 1.6796875, |
| "learning_rate": 1.0025835437160523e-05, |
| "loss": 2.1363, |
| "mean_token_accuracy": 0.7570828595602909, |
| "mean_token_accuracy_utility": 0.675681160673971, |
| "step": 1485 |
| }, |
| { |
| "epoch": 0.5512393636699963, |
| "grad_norm": 1.6796875, |
| "learning_rate": 9.96124689814801e-06, |
| "loss": 1.8993, |
| "mean_token_accuracy": 0.7508178948059105, |
| "mean_token_accuracy_utility": 0.7173593976782351, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.5530891601923789, |
| "grad_norm": 1.8203125, |
| "learning_rate": 9.89665997579133e-06, |
| "loss": 1.9202, |
| "mean_token_accuracy": 0.7546300337087654, |
| "mean_token_accuracy_utility": 0.709141543826316, |
| "step": 1495 |
| }, |
| { |
| "epoch": 0.5549389567147613, |
| "grad_norm": 1.7578125, |
| "learning_rate": 9.832077364450915e-06, |
| "loss": 1.8814, |
| "mean_token_accuracy": 0.7796635355617355, |
| "mean_token_accuracy_utility": 0.7012997207582459, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.5549389567147613, |
| "eval_mean_token_accuracy": 0.752907188347999, |
| "eval_reasoning_loss": 0.8318967819213867, |
| "eval_reasoning_runtime": 29.3383, |
| "eval_reasoning_samples_per_second": 4.397, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.5549389567147613, |
| "eval_mean_token_accuracy": 0.7010242591932953, |
| "eval_utility_loss": 1.1551748514175415, |
| "eval_utility_runtime": 11.0753, |
| "eval_utility_samples_per_second": 4.424, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.5567887532371439, |
| "grad_norm": 1.7578125, |
| "learning_rate": 9.767501758307376e-06, |
| "loss": 1.8664, |
| "mean_token_accuracy": 0.7678236622457002, |
| "mean_token_accuracy_utility": 0.7097331362582628, |
| "step": 1505 |
| }, |
| { |
| "epoch": 0.5586385497595264, |
| "grad_norm": 1.8515625, |
| "learning_rate": 9.70293585124907e-06, |
| "loss": 1.8541, |
| "mean_token_accuracy": 0.77482585397328, |
| "mean_token_accuracy_utility": 0.7118054725744221, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.560488346281909, |
| "grad_norm": 1.9140625, |
| "learning_rate": 9.638382336759749e-06, |
| "loss": 1.8747, |
| "mean_token_accuracy": 0.7582848524859117, |
| "mean_token_accuracy_utility": 0.7178581143968701, |
| "step": 1515 |
| }, |
| { |
| "epoch": 0.5623381428042915, |
| "grad_norm": 1.90625, |
| "learning_rate": 9.57384390780618e-06, |
| "loss": 1.8974, |
| "mean_token_accuracy": 0.7656608138178433, |
| "mean_token_accuracy_utility": 0.706474688962956, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.5641879393266741, |
| "grad_norm": 1.734375, |
| "learning_rate": 9.50932325672582e-06, |
| "loss": 1.9503, |
| "mean_token_accuracy": 0.7511089796337844, |
| "mean_token_accuracy_utility": 0.7054385580057257, |
| "step": 1525 |
| }, |
| { |
| "epoch": 0.5660377358490566, |
| "grad_norm": 1.6484375, |
| "learning_rate": 9.444823075114481e-06, |
| "loss": 1.8998, |
| "mean_token_accuracy": 0.7649870820548953, |
| "mean_token_accuracy_utility": 0.7077748060228023, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.5678875323714392, |
| "grad_norm": 1.6875, |
| "learning_rate": 9.380346053714055e-06, |
| "loss": 1.8363, |
| "mean_token_accuracy": 0.7753260220241405, |
| "mean_token_accuracy_utility": 0.7105223771562822, |
| "step": 1535 |
| }, |
| { |
| "epoch": 0.5697373288938217, |
| "grad_norm": 1.640625, |
| "learning_rate": 9.315894882300269e-06, |
| "loss": 1.8147, |
| "mean_token_accuracy": 0.7828657909013241, |
| "mean_token_accuracy_utility": 0.7119973971520126, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.5715871254162043, |
| "grad_norm": 1.796875, |
| "learning_rate": 9.251472249570458e-06, |
| "loss": 1.8696, |
| "mean_token_accuracy": 0.759969374619339, |
| "mean_token_accuracy_utility": 0.7162541801192138, |
| "step": 1545 |
| }, |
| { |
| "epoch": 0.5734369219385868, |
| "grad_norm": 1.671875, |
| "learning_rate": 9.187080843031436e-06, |
| "loss": 1.8676, |
| "mean_token_accuracy": 0.7688888370901221, |
| "mean_token_accuracy_utility": 0.7104964134933971, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.5752867184609693, |
| "grad_norm": 1.765625, |
| "learning_rate": 9.12272334888734e-06, |
| "loss": 1.941, |
| "mean_token_accuracy": 0.7553321782142721, |
| "mean_token_accuracy_utility": 0.7055089646448434, |
| "step": 1555 |
| }, |
| { |
| "epoch": 0.5771365149833518, |
| "grad_norm": 1.59375, |
| "learning_rate": 9.058402451927599e-06, |
| "loss": 1.8837, |
| "mean_token_accuracy": 0.7720471983455863, |
| "mean_token_accuracy_utility": 0.7056620628956245, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.5789863115057343, |
| "grad_norm": 1.7421875, |
| "learning_rate": 8.99412083541492e-06, |
| "loss": 1.8297, |
| "mean_token_accuracy": 0.7674895132721338, |
| "mean_token_accuracy_utility": 0.7188344987218837, |
| "step": 1565 |
| }, |
| { |
| "epoch": 0.5808361080281169, |
| "grad_norm": 1.7890625, |
| "learning_rate": 8.929881180973365e-06, |
| "loss": 1.8946, |
| "mean_token_accuracy": 0.7574488637326489, |
| "mean_token_accuracy_utility": 0.7150735543607184, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.5826859045504994, |
| "grad_norm": 1.671875, |
| "learning_rate": 8.865686168476458e-06, |
| "loss": 1.9784, |
| "mean_token_accuracy": 0.7610113876278213, |
| "mean_token_accuracy_utility": 0.6954516822148624, |
| "step": 1575 |
| }, |
| { |
| "epoch": 0.584535701072882, |
| "grad_norm": 1.6953125, |
| "learning_rate": 8.801538475935413e-06, |
| "loss": 1.8537, |
| "mean_token_accuracy": 0.770962501928553, |
| "mean_token_accuracy_utility": 0.7113820398760473, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.5863854975952645, |
| "grad_norm": 1.6953125, |
| "learning_rate": 8.737440779387398e-06, |
| "loss": 1.9346, |
| "mean_token_accuracy": 0.7605264131370886, |
| "mean_token_accuracy_utility": 0.7042874293585664, |
| "step": 1585 |
| }, |
| { |
| "epoch": 0.5882352941176471, |
| "grad_norm": 1.7109375, |
| "learning_rate": 8.673395752783924e-06, |
| "loss": 1.9944, |
| "mean_token_accuracy": 0.7595051958816732, |
| "mean_token_accuracy_utility": 0.6915728947127879, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.5900850906400296, |
| "grad_norm": 1.734375, |
| "learning_rate": 8.609406067879258e-06, |
| "loss": 1.8828, |
| "mean_token_accuracy": 0.7675511781513877, |
| "mean_token_accuracy_utility": 0.7095317684209844, |
| "step": 1595 |
| }, |
| { |
| "epoch": 0.5919348871624122, |
| "grad_norm": 1.65625, |
| "learning_rate": 8.54547439411901e-06, |
| "loss": 1.9532, |
| "mean_token_accuracy": 0.7566409413214015, |
| "mean_token_accuracy_utility": 0.7046242016842783, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.5919348871624122, |
| "eval_mean_token_accuracy": 0.753047889296913, |
| "eval_reasoning_loss": 0.8311854004859924, |
| "eval_reasoning_runtime": 29.3007, |
| "eval_reasoning_samples_per_second": 4.403, |
| "eval_reasoning_steps_per_second": 1.126, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.5919348871624122, |
| "eval_mean_token_accuracy": 0.7011958851068323, |
| "eval_utility_loss": 1.1550215482711792, |
| "eval_utility_runtime": 11.065, |
| "eval_utility_samples_per_second": 4.428, |
| "eval_utility_steps_per_second": 1.175, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.5937846836847946, |
| "grad_norm": 1.6796875, |
| "learning_rate": 8.481603398528727e-06, |
| "loss": 1.9119, |
| "mean_token_accuracy": 0.7770164233730589, |
| "mean_token_accuracy_utility": 0.6982484597512488, |
| "step": 1605 |
| }, |
| { |
| "epoch": 0.5956344802071772, |
| "grad_norm": 1.875, |
| "learning_rate": 8.417795745602679e-06, |
| "loss": 1.8655, |
| "mean_token_accuracy": 0.7596301730356192, |
| "mean_token_accuracy_utility": 0.7171135496494735, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.5974842767295597, |
| "grad_norm": 1.6484375, |
| "learning_rate": 8.35405409719266e-06, |
| "loss": 1.9374, |
| "mean_token_accuracy": 0.7553295564651183, |
| "mean_token_accuracy_utility": 0.7039287028432685, |
| "step": 1615 |
| }, |
| { |
| "epoch": 0.5993340732519423, |
| "grad_norm": 1.7421875, |
| "learning_rate": 8.290381112396989e-06, |
| "loss": 1.8777, |
| "mean_token_accuracy": 0.7686658682924972, |
| "mean_token_accuracy_utility": 0.7099313172363587, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.6011838697743248, |
| "grad_norm": 1.671875, |
| "learning_rate": 8.226779447449538e-06, |
| "loss": 1.8631, |
| "mean_token_accuracy": 0.7704549232369117, |
| "mean_token_accuracy_utility": 0.7094036644170221, |
| "step": 1625 |
| }, |
| { |
| "epoch": 0.6030336662967074, |
| "grad_norm": 1.7109375, |
| "learning_rate": 8.163251755608954e-06, |
| "loss": 1.9606, |
| "mean_token_accuracy": 0.7632959412844009, |
| "mean_token_accuracy_utility": 0.6984329143208564, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.6048834628190899, |
| "grad_norm": 1.71875, |
| "learning_rate": 8.099800687047958e-06, |
| "loss": 1.9155, |
| "mean_token_accuracy": 0.7656983548723187, |
| "mean_token_accuracy_utility": 0.7040221827613207, |
| "step": 1635 |
| }, |
| { |
| "epoch": 0.6067332593414725, |
| "grad_norm": 1.734375, |
| "learning_rate": 8.036428888742789e-06, |
| "loss": 1.872, |
| "mean_token_accuracy": 0.7635138610838019, |
| "mean_token_accuracy_utility": 0.7134666086054345, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.608583055863855, |
| "grad_norm": 1.78125, |
| "learning_rate": 7.973139004362794e-06, |
| "loss": 1.9227, |
| "mean_token_accuracy": 0.7618785872104308, |
| "mean_token_accuracy_utility": 0.7120191163693543, |
| "step": 1645 |
| }, |
| { |
| "epoch": 0.6104328523862376, |
| "grad_norm": 1.6640625, |
| "learning_rate": 7.909933674160118e-06, |
| "loss": 1.9171, |
| "mean_token_accuracy": 0.7553160632099373, |
| "mean_token_accuracy_utility": 0.7091542455961231, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.61228264890862, |
| "grad_norm": 1.7265625, |
| "learning_rate": 7.846815534859592e-06, |
| "loss": 1.9655, |
| "mean_token_accuracy": 0.7612304455947001, |
| "mean_token_accuracy_utility": 0.7008781679122252, |
| "step": 1655 |
| }, |
| { |
| "epoch": 0.6141324454310025, |
| "grad_norm": 1.625, |
| "learning_rate": 7.7837872195487e-06, |
| "loss": 1.8382, |
| "mean_token_accuracy": 0.762500115726061, |
| "mean_token_accuracy_utility": 0.720514731046623, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.6159822419533851, |
| "grad_norm": 1.6953125, |
| "learning_rate": 7.720851357567778e-06, |
| "loss": 1.8921, |
| "mean_token_accuracy": 0.7705835757724142, |
| "mean_token_accuracy_utility": 0.7043785335608436, |
| "step": 1665 |
| }, |
| { |
| "epoch": 0.6178320384757676, |
| "grad_norm": 1.7578125, |
| "learning_rate": 7.658010574400285e-06, |
| "loss": 1.9357, |
| "mean_token_accuracy": 0.7660957690014132, |
| "mean_token_accuracy_utility": 0.7040110323657793, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.6196818349981502, |
| "grad_norm": 1.75, |
| "learning_rate": 7.595267491563311e-06, |
| "loss": 1.8873, |
| "mean_token_accuracy": 0.7660623884182817, |
| "mean_token_accuracy_utility": 0.7073874089752674, |
| "step": 1675 |
| }, |
| { |
| "epoch": 0.6215316315205327, |
| "grad_norm": 1.828125, |
| "learning_rate": 7.532624726498181e-06, |
| "loss": 1.9384, |
| "mean_token_accuracy": 0.7628961474448933, |
| "mean_token_accuracy_utility": 0.7014302035606499, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.6233814280429153, |
| "grad_norm": 1.7109375, |
| "learning_rate": 7.470084892461305e-06, |
| "loss": 1.9224, |
| "mean_token_accuracy": 0.7562725815233886, |
| "mean_token_accuracy_utility": 0.7091136372505187, |
| "step": 1685 |
| }, |
| { |
| "epoch": 0.6252312245652978, |
| "grad_norm": 1.625, |
| "learning_rate": 7.407650598415119e-06, |
| "loss": 1.8787, |
| "mean_token_accuracy": 0.7598822565195403, |
| "mean_token_accuracy_utility": 0.7160148789321121, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.6270810210876804, |
| "grad_norm": 1.65625, |
| "learning_rate": 7.34532444891928e-06, |
| "loss": 1.9218, |
| "mean_token_accuracy": 0.7641593221873821, |
| "mean_token_accuracy_utility": 0.7054986663701813, |
| "step": 1695 |
| }, |
| { |
| "epoch": 0.6289308176100629, |
| "grad_norm": 1.6875, |
| "learning_rate": 7.283109044021987e-06, |
| "loss": 2.065, |
| "mean_token_accuracy": 0.7413078086943707, |
| "mean_token_accuracy_utility": 0.6927783837753865, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.6289308176100629, |
| "eval_mean_token_accuracy": 0.7530756402459806, |
| "eval_reasoning_loss": 0.8308614492416382, |
| "eval_reasoning_runtime": 29.343, |
| "eval_reasoning_samples_per_second": 4.396, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.6289308176100629, |
| "eval_mean_token_accuracy": 0.7013199259866364, |
| "eval_utility_loss": 1.1547558307647705, |
| "eval_utility_runtime": 11.0815, |
| "eval_utility_samples_per_second": 4.422, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.6307806141324455, |
| "grad_norm": 1.7578125, |
| "learning_rate": 7.221006979151546e-06, |
| "loss": 1.9321, |
| "mean_token_accuracy": 0.7605704064503289, |
| "mean_token_accuracy_utility": 0.7091791189032329, |
| "step": 1705 |
| }, |
| { |
| "epoch": 0.632630410654828, |
| "grad_norm": 1.828125, |
| "learning_rate": 7.159020845008056e-06, |
| "loss": 2.0318, |
| "mean_token_accuracy": 0.7409577187854014, |
| "mean_token_accuracy_utility": 0.6968604364614096, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.6344802071772105, |
| "grad_norm": 1.671875, |
| "learning_rate": 7.097153227455379e-06, |
| "loss": 1.9028, |
| "mean_token_accuracy": 0.7630189231670592, |
| "mean_token_accuracy_utility": 0.7057549276990821, |
| "step": 1715 |
| }, |
| { |
| "epoch": 0.636330003699593, |
| "grad_norm": 1.6953125, |
| "learning_rate": 7.035406707413221e-06, |
| "loss": 1.9569, |
| "mean_token_accuracy": 0.7530720329961921, |
| "mean_token_accuracy_utility": 0.7058222693084728, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.6381798002219756, |
| "grad_norm": 1.640625, |
| "learning_rate": 6.973783860749499e-06, |
| "loss": 1.9082, |
| "mean_token_accuracy": 0.759830516957526, |
| "mean_token_accuracy_utility": 0.7086231346311203, |
| "step": 1725 |
| }, |
| { |
| "epoch": 0.6400295967443581, |
| "grad_norm": 1.5703125, |
| "learning_rate": 6.912287258172869e-06, |
| "loss": 1.8185, |
| "mean_token_accuracy": 0.7833409104069599, |
| "mean_token_accuracy_utility": 0.708338157842777, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.6418793932667407, |
| "grad_norm": 1.6953125, |
| "learning_rate": 6.8509194651254825e-06, |
| "loss": 2.0195, |
| "mean_token_accuracy": 0.7546006433692694, |
| "mean_token_accuracy_utility": 0.6950483440962018, |
| "step": 1735 |
| }, |
| { |
| "epoch": 0.6437291897891232, |
| "grad_norm": 1.7421875, |
| "learning_rate": 6.789683041675977e-06, |
| "loss": 1.9224, |
| "mean_token_accuracy": 0.7562899798586009, |
| "mean_token_accuracy_utility": 0.7097862169399906, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.6455789863115058, |
| "grad_norm": 1.7578125, |
| "learning_rate": 6.7285805424126585e-06, |
| "loss": 1.8884, |
| "mean_token_accuracy": 0.7537024629134665, |
| "mean_token_accuracy_utility": 0.7177109690253828, |
| "step": 1745 |
| }, |
| { |
| "epoch": 0.6474287828338883, |
| "grad_norm": 1.6796875, |
| "learning_rate": 6.667614516336949e-06, |
| "loss": 1.9028, |
| "mean_token_accuracy": 0.7623305677935669, |
| "mean_token_accuracy_utility": 0.7128328714097937, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.6492785793562708, |
| "grad_norm": 1.671875, |
| "learning_rate": 6.60678750675704e-06, |
| "loss": 1.9264, |
| "mean_token_accuracy": 0.7531829418415963, |
| "mean_token_accuracy_utility": 0.7108618299273295, |
| "step": 1755 |
| }, |
| { |
| "epoch": 0.6511283758786534, |
| "grad_norm": 1.6328125, |
| "learning_rate": 6.546102051181808e-06, |
| "loss": 1.9506, |
| "mean_token_accuracy": 0.7762314769699459, |
| "mean_token_accuracy_utility": 0.6958801158142984, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.6529781724010358, |
| "grad_norm": 1.75, |
| "learning_rate": 6.485560681214933e-06, |
| "loss": 1.9458, |
| "mean_token_accuracy": 0.757005586028672, |
| "mean_token_accuracy_utility": 0.7056531742230259, |
| "step": 1765 |
| }, |
| { |
| "epoch": 0.6548279689234184, |
| "grad_norm": 1.703125, |
| "learning_rate": 6.425165922449319e-06, |
| "loss": 1.942, |
| "mean_token_accuracy": 0.750982745619041, |
| "mean_token_accuracy_utility": 0.7074444968206576, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.6566777654458009, |
| "grad_norm": 1.8046875, |
| "learning_rate": 6.364920294361701e-06, |
| "loss": 1.8688, |
| "mean_token_accuracy": 0.7608773375651332, |
| "mean_token_accuracy_utility": 0.718894096195611, |
| "step": 1775 |
| }, |
| { |
| "epoch": 0.6585275619681835, |
| "grad_norm": 1.7578125, |
| "learning_rate": 6.3048263102075766e-06, |
| "loss": 1.9169, |
| "mean_token_accuracy": 0.7583410260303272, |
| "mean_token_accuracy_utility": 0.7082648240275172, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.660377358490566, |
| "grad_norm": 1.65625, |
| "learning_rate": 6.244886476916325e-06, |
| "loss": 1.9275, |
| "mean_token_accuracy": 0.7587991089424289, |
| "mean_token_accuracy_utility": 0.7068662597024867, |
| "step": 1785 |
| }, |
| { |
| "epoch": 0.6622271550129486, |
| "grad_norm": 1.75, |
| "learning_rate": 6.185103294986654e-06, |
| "loss": 2.0147, |
| "mean_token_accuracy": 0.7621971879818872, |
| "mean_token_accuracy_utility": 0.6944179611177244, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.6640769515353311, |
| "grad_norm": 1.734375, |
| "learning_rate": 6.125479258382268e-06, |
| "loss": 1.9538, |
| "mean_token_accuracy": 0.7610684257169326, |
| "mean_token_accuracy_utility": 0.6958711064116387, |
| "step": 1795 |
| }, |
| { |
| "epoch": 0.6659267480577137, |
| "grad_norm": 1.8203125, |
| "learning_rate": 6.066016854427846e-06, |
| "loss": 1.8437, |
| "mean_token_accuracy": 0.763900988911854, |
| "mean_token_accuracy_utility": 0.7202450089192622, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.6659267480577137, |
| "eval_mean_token_accuracy": 0.7530941422944204, |
| "eval_reasoning_loss": 0.8304892778396606, |
| "eval_reasoning_runtime": 29.3487, |
| "eval_reasoning_samples_per_second": 4.395, |
| "eval_reasoning_steps_per_second": 1.124, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.6659267480577137, |
| "eval_mean_token_accuracy": 0.7015515508565339, |
| "eval_utility_loss": 1.154640793800354, |
| "eval_utility_runtime": 11.0706, |
| "eval_utility_samples_per_second": 4.426, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.6677765445800962, |
| "grad_norm": 1.953125, |
| "learning_rate": 6.006718563705258e-06, |
| "loss": 1.918, |
| "mean_token_accuracy": 0.7524923691819041, |
| "mean_token_accuracy_utility": 0.7116922840643984, |
| "step": 1805 |
| }, |
| { |
| "epoch": 0.6696263411024788, |
| "grad_norm": 1.703125, |
| "learning_rate": 5.947586859950103e-06, |
| "loss": 1.8249, |
| "mean_token_accuracy": 0.7699044543610911, |
| "mean_token_accuracy_utility": 0.7215691870653829, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.6714761376248612, |
| "grad_norm": 1.8359375, |
| "learning_rate": 5.888624209948495e-06, |
| "loss": 2.026, |
| "mean_token_accuracy": 0.7620448071187224, |
| "mean_token_accuracy_utility": 0.6918027243571019, |
| "step": 1815 |
| }, |
| { |
| "epoch": 0.6733259341472438, |
| "grad_norm": 1.765625, |
| "learning_rate": 5.82983307343416e-06, |
| "loss": 1.8826, |
| "mean_token_accuracy": 0.7719674603568827, |
| "mean_token_accuracy_utility": 0.7058305906252913, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.6751757306696263, |
| "grad_norm": 1.9296875, |
| "learning_rate": 5.771215902985848e-06, |
| "loss": 1.8901, |
| "mean_token_accuracy": 0.7670450884420219, |
| "mean_token_accuracy_utility": 0.7062487643063677, |
| "step": 1825 |
| }, |
| { |
| "epoch": 0.6770255271920089, |
| "grad_norm": 1.71875, |
| "learning_rate": 5.712775143924978e-06, |
| "loss": 1.9449, |
| "mean_token_accuracy": 0.7700723011119862, |
| "mean_token_accuracy_utility": 0.6943705642251716, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.6788753237143914, |
| "grad_norm": 1.6484375, |
| "learning_rate": 5.654513234213663e-06, |
| "loss": 1.8855, |
| "mean_token_accuracy": 0.7636314672503743, |
| "mean_token_accuracy_utility": 0.7141322120906304, |
| "step": 1835 |
| }, |
| { |
| "epoch": 0.680725120236774, |
| "grad_norm": 1.96875, |
| "learning_rate": 5.596432604352982e-06, |
| "loss": 1.9735, |
| "mean_token_accuracy": 0.7696641527632032, |
| "mean_token_accuracy_utility": 0.6897577450205975, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.6825749167591565, |
| "grad_norm": 1.640625, |
| "learning_rate": 5.538535677281608e-06, |
| "loss": 1.9799, |
| "mean_token_accuracy": 0.7586655058937859, |
| "mean_token_accuracy_utility": 0.6967709029756539, |
| "step": 1845 |
| }, |
| { |
| "epoch": 0.684424713281539, |
| "grad_norm": 1.8046875, |
| "learning_rate": 5.480824868274699e-06, |
| "loss": 1.9445, |
| "mean_token_accuracy": 0.7592704182281267, |
| "mean_token_accuracy_utility": 0.706764977296716, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.6862745098039216, |
| "grad_norm": 1.7421875, |
| "learning_rate": 5.423302584843186e-06, |
| "loss": 2.0121, |
| "mean_token_accuracy": 0.7524013767420913, |
| "mean_token_accuracy_utility": 0.6967962315818685, |
| "step": 1855 |
| }, |
| { |
| "epoch": 0.6881243063263041, |
| "grad_norm": 1.7421875, |
| "learning_rate": 5.365971226633294e-06, |
| "loss": 1.8889, |
| "mean_token_accuracy": 0.7736774001058592, |
| "mean_token_accuracy_utility": 0.7032134873445884, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.6899741028486867, |
| "grad_norm": 1.9453125, |
| "learning_rate": 5.308833185326472e-06, |
| "loss": 1.9991, |
| "mean_token_accuracy": 0.7587936979751736, |
| "mean_token_accuracy_utility": 0.6969532070833659, |
| "step": 1865 |
| }, |
| { |
| "epoch": 0.6918238993710691, |
| "grad_norm": 1.7890625, |
| "learning_rate": 5.25189084453959e-06, |
| "loss": 1.9499, |
| "mean_token_accuracy": 0.7622456209101666, |
| "mean_token_accuracy_utility": 0.6961910741051796, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.6936736958934517, |
| "grad_norm": 1.6953125, |
| "learning_rate": 5.19514657972553e-06, |
| "loss": 1.8592, |
| "mean_token_accuracy": 0.7784553507444147, |
| "mean_token_accuracy_utility": 0.7103179638317741, |
| "step": 1875 |
| }, |
| { |
| "epoch": 0.6955234924158342, |
| "grad_norm": 1.59375, |
| "learning_rate": 5.138602758074062e-06, |
| "loss": 1.832, |
| "mean_token_accuracy": 0.7777301685588347, |
| "mean_token_accuracy_utility": 0.7097199365417791, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.6973732889382168, |
| "grad_norm": 1.8984375, |
| "learning_rate": 5.082261738413124e-06, |
| "loss": 1.9694, |
| "mean_token_accuracy": 0.7489460613979676, |
| "mean_token_accuracy_utility": 0.7052067742244473, |
| "step": 1885 |
| }, |
| { |
| "epoch": 0.6992230854605993, |
| "grad_norm": 1.6875, |
| "learning_rate": 5.026125871110383e-06, |
| "loss": 1.9708, |
| "mean_token_accuracy": 0.7619654595785151, |
| "mean_token_accuracy_utility": 0.6974978879848625, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.7010728819829819, |
| "grad_norm": 1.703125, |
| "learning_rate": 4.970197497975216e-06, |
| "loss": 1.9318, |
| "mean_token_accuracy": 0.7723678642403946, |
| "mean_token_accuracy_utility": 0.6929367847930574, |
| "step": 1895 |
| }, |
| { |
| "epoch": 0.7029226785053644, |
| "grad_norm": 1.578125, |
| "learning_rate": 4.914478952160997e-06, |
| "loss": 1.9382, |
| "mean_token_accuracy": 0.7568420239371472, |
| "mean_token_accuracy_utility": 0.7113332794348479, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.7029226785053644, |
| "eval_mean_token_accuracy": 0.7532240204785101, |
| "eval_reasoning_loss": 0.830326497554779, |
| "eval_reasoning_runtime": 29.3292, |
| "eval_reasoning_samples_per_second": 4.398, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.7029226785053644, |
| "eval_mean_token_accuracy": 0.7015841190182154, |
| "eval_utility_loss": 1.1547009944915771, |
| "eval_utility_runtime": 11.0687, |
| "eval_utility_samples_per_second": 4.427, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.704772475027747, |
| "grad_norm": 1.671875, |
| "learning_rate": 4.858972558067784e-06, |
| "loss": 1.8339, |
| "mean_token_accuracy": 0.7803520899198514, |
| "mean_token_accuracy_utility": 0.7097957780084616, |
| "step": 1905 |
| }, |
| { |
| "epoch": 0.7066222715501295, |
| "grad_norm": 1.6484375, |
| "learning_rate": 4.803680631245333e-06, |
| "loss": 1.8022, |
| "mean_token_accuracy": 0.7756003885630276, |
| "mean_token_accuracy_utility": 0.7218335015100162, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.7084720680725121, |
| "grad_norm": 1.703125, |
| "learning_rate": 4.748605478296508e-06, |
| "loss": 1.8753, |
| "mean_token_accuracy": 0.7735854369750307, |
| "mean_token_accuracy_utility": 0.7050308189123164, |
| "step": 1915 |
| }, |
| { |
| "epoch": 0.7103218645948945, |
| "grad_norm": 1.859375, |
| "learning_rate": 4.6937493967810665e-06, |
| "loss": 1.9802, |
| "mean_token_accuracy": 0.7679639318052438, |
| "mean_token_accuracy_utility": 0.691909069916302, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.7121716611172771, |
| "grad_norm": 1.7734375, |
| "learning_rate": 4.639114675119797e-06, |
| "loss": 1.99, |
| "mean_token_accuracy": 0.7546679135737763, |
| "mean_token_accuracy_utility": 0.6977523046256351, |
| "step": 1925 |
| }, |
| { |
| "epoch": 0.7140214576396596, |
| "grad_norm": 1.75, |
| "learning_rate": 4.584703592499068e-06, |
| "loss": 1.9333, |
| "mean_token_accuracy": 0.7626703560220914, |
| "mean_token_accuracy_utility": 0.7033857944103676, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.7158712541620422, |
| "grad_norm": 1.6171875, |
| "learning_rate": 4.530518418775734e-06, |
| "loss": 1.895, |
| "mean_token_accuracy": 0.7656944301349773, |
| "mean_token_accuracy_utility": 0.7083699338219586, |
| "step": 1935 |
| }, |
| { |
| "epoch": 0.7177210506844247, |
| "grad_norm": 1.6171875, |
| "learning_rate": 4.476561414382451e-06, |
| "loss": 1.7729, |
| "mean_token_accuracy": 0.7816831029624717, |
| "mean_token_accuracy_utility": 0.7223184379420882, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.7195708472068072, |
| "grad_norm": 1.8984375, |
| "learning_rate": 4.422834830233378e-06, |
| "loss": 1.8642, |
| "mean_token_accuracy": 0.7655339244747723, |
| "mean_token_accuracy_utility": 0.7148753998071007, |
| "step": 1945 |
| }, |
| { |
| "epoch": 0.7214206437291898, |
| "grad_norm": 1.9375, |
| "learning_rate": 4.369340907630284e-06, |
| "loss": 1.949, |
| "mean_token_accuracy": 0.7529812988253687, |
| "mean_token_accuracy_utility": 0.7055938470732839, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.7232704402515723, |
| "grad_norm": 1.640625, |
| "learning_rate": 4.3160818781690286e-06, |
| "loss": 1.9039, |
| "mean_token_accuracy": 0.7586304410613351, |
| "mean_token_accuracy_utility": 0.7125044183051326, |
| "step": 1955 |
| }, |
| { |
| "epoch": 0.7251202367739549, |
| "grad_norm": 1.7421875, |
| "learning_rate": 4.26305996364649e-06, |
| "loss": 1.9304, |
| "mean_token_accuracy": 0.7669084693034954, |
| "mean_token_accuracy_utility": 0.7008601714887861, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.7269700332963374, |
| "grad_norm": 1.7265625, |
| "learning_rate": 4.210277375967855e-06, |
| "loss": 1.8599, |
| "mean_token_accuracy": 0.7608192294490922, |
| "mean_token_accuracy_utility": 0.7162971935262964, |
| "step": 1965 |
| }, |
| { |
| "epoch": 0.72881982981872, |
| "grad_norm": 2.125, |
| "learning_rate": 4.157736317054373e-06, |
| "loss": 2.0052, |
| "mean_token_accuracy": 0.7597422414759062, |
| "mean_token_accuracy_utility": 0.6915794828628743, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.7306696263411024, |
| "grad_norm": 1.6328125, |
| "learning_rate": 4.105438978751465e-06, |
| "loss": 1.9301, |
| "mean_token_accuracy": 0.7665287715133944, |
| "mean_token_accuracy_utility": 0.6972798186518822, |
| "step": 1975 |
| }, |
| { |
| "epoch": 0.732519422863485, |
| "grad_norm": 1.8515625, |
| "learning_rate": 4.053387542737323e-06, |
| "loss": 1.8781, |
| "mean_token_accuracy": 0.7681039899190912, |
| "mean_token_accuracy_utility": 0.7134526722121814, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.7343692193858675, |
| "grad_norm": 1.71875, |
| "learning_rate": 4.001584180431869e-06, |
| "loss": 1.9248, |
| "mean_token_accuracy": 0.7725860169247134, |
| "mean_token_accuracy_utility": 0.6991295415296125, |
| "step": 1985 |
| }, |
| { |
| "epoch": 0.7362190159082501, |
| "grad_norm": 1.7109375, |
| "learning_rate": 3.950031052906183e-06, |
| "loss": 1.9266, |
| "mean_token_accuracy": 0.7670044405340277, |
| "mean_token_accuracy_utility": 0.7031292204736159, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.7380688124306326, |
| "grad_norm": 1.8125, |
| "learning_rate": 3.898730310792346e-06, |
| "loss": 1.9224, |
| "mean_token_accuracy": 0.7581986720530913, |
| "mean_token_accuracy_utility": 0.7114034306375081, |
| "step": 1995 |
| }, |
| { |
| "epoch": 0.7399186089530152, |
| "grad_norm": 1.7578125, |
| "learning_rate": 3.847684094193733e-06, |
| "loss": 1.9109, |
| "mean_token_accuracy": 0.7541041568892453, |
| "mean_token_accuracy_utility": 0.7099421449410203, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.7399186089530152, |
| "eval_mean_token_accuracy": 0.753221844406543, |
| "eval_reasoning_loss": 0.8301464915275574, |
| "eval_reasoning_runtime": 29.3115, |
| "eval_reasoning_samples_per_second": 4.401, |
| "eval_reasoning_steps_per_second": 1.126, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.7399186089530152, |
| "eval_mean_token_accuracy": 0.7012614251187116, |
| "eval_utility_loss": 1.1547011137008667, |
| "eval_utility_runtime": 11.085, |
| "eval_utility_samples_per_second": 4.42, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.7417684054753977, |
| "grad_norm": 1.7734375, |
| "learning_rate": 3.7968945325957175e-06, |
| "loss": 2.0007, |
| "mean_token_accuracy": 0.7541255533024882, |
| "mean_token_accuracy_utility": 0.6937630198102139, |
| "step": 2005 |
| }, |
| { |
| "epoch": 0.7436182019977803, |
| "grad_norm": 1.9140625, |
| "learning_rate": 3.746363744776844e-06, |
| "loss": 1.9707, |
| "mean_token_accuracy": 0.7538567711761918, |
| "mean_token_accuracy_utility": 0.6986743048232928, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.7454679985201628, |
| "grad_norm": 1.7265625, |
| "learning_rate": 3.69609383872045e-06, |
| "loss": 1.9235, |
| "mean_token_accuracy": 0.7634287052197951, |
| "mean_token_accuracy_utility": 0.7016636518758744, |
| "step": 2015 |
| }, |
| { |
| "epoch": 0.7473177950425454, |
| "grad_norm": 1.65625, |
| "learning_rate": 3.646086911526706e-06, |
| "loss": 1.8186, |
| "mean_token_accuracy": 0.7717394511549277, |
| "mean_token_accuracy_utility": 0.7169230836613947, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.7491675915649278, |
| "grad_norm": 1.7265625, |
| "learning_rate": 3.5963450493251552e-06, |
| "loss": 1.8244, |
| "mean_token_accuracy": 0.757721043672541, |
| "mean_token_accuracy_utility": 0.7301172844428198, |
| "step": 2025 |
| }, |
| { |
| "epoch": 0.7510173880873104, |
| "grad_norm": 1.703125, |
| "learning_rate": 3.5468703271876626e-06, |
| "loss": 1.8702, |
| "mean_token_accuracy": 0.7712568652035663, |
| "mean_token_accuracy_utility": 0.7098278135896063, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.7528671846096929, |
| "grad_norm": 1.875, |
| "learning_rate": 3.4976648090418685e-06, |
| "loss": 1.942, |
| "mean_token_accuracy": 0.7641528888868313, |
| "mean_token_accuracy_utility": 0.7013159445125456, |
| "step": 2035 |
| }, |
| { |
| "epoch": 0.7547169811320755, |
| "grad_norm": 1.9453125, |
| "learning_rate": 3.4487305475850754e-06, |
| "loss": 2.0093, |
| "mean_token_accuracy": 0.7577362497529686, |
| "mean_token_accuracy_utility": 0.6956651922671905, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.756566777654458, |
| "grad_norm": 1.6640625, |
| "learning_rate": 3.400069584198633e-06, |
| "loss": 1.9926, |
| "mean_token_accuracy": 0.7626071235216366, |
| "mean_token_accuracy_utility": 0.6933814029100341, |
| "step": 2045 |
| }, |
| { |
| "epoch": 0.7584165741768405, |
| "grad_norm": 1.8125, |
| "learning_rate": 3.351683948862753e-06, |
| "loss": 1.9443, |
| "mean_token_accuracy": 0.7548732633422665, |
| "mean_token_accuracy_utility": 0.6999794101062377, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.7602663706992231, |
| "grad_norm": 1.8984375, |
| "learning_rate": 3.3035756600718515e-06, |
| "loss": 1.9229, |
| "mean_token_accuracy": 0.7659286348479356, |
| "mean_token_accuracy_utility": 0.7010594767378565, |
| "step": 2055 |
| }, |
| { |
| "epoch": 0.7621161672216056, |
| "grad_norm": 1.8671875, |
| "learning_rate": 3.2557467247503205e-06, |
| "loss": 1.9161, |
| "mean_token_accuracy": 0.762718426217766, |
| "mean_token_accuracy_utility": 0.7106908637528029, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.7639659637439882, |
| "grad_norm": 1.6875, |
| "learning_rate": 3.208199138168826e-06, |
| "loss": 1.9642, |
| "mean_token_accuracy": 0.7501900846314478, |
| "mean_token_accuracy_utility": 0.7093633700885851, |
| "step": 2065 |
| }, |
| { |
| "epoch": 0.7658157602663707, |
| "grad_norm": 1.6796875, |
| "learning_rate": 3.160934883861053e-06, |
| "loss": 1.904, |
| "mean_token_accuracy": 0.7574337606238709, |
| "mean_token_accuracy_utility": 0.7125443972112077, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.7676655567887533, |
| "grad_norm": 1.6953125, |
| "learning_rate": 3.113955933540973e-06, |
| "loss": 1.8742, |
| "mean_token_accuracy": 0.7839844545347672, |
| "mean_token_accuracy_utility": 0.7015719934361019, |
| "step": 2075 |
| }, |
| { |
| "epoch": 0.7695153533111357, |
| "grad_norm": 1.75, |
| "learning_rate": 3.0672642470205794e-06, |
| "loss": 2.0582, |
| "mean_token_accuracy": 0.7615933941717115, |
| "mean_token_accuracy_utility": 0.6855634578885514, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.7713651498335183, |
| "grad_norm": 1.84375, |
| "learning_rate": 3.020861772128145e-06, |
| "loss": 1.962, |
| "mean_token_accuracy": 0.7471739866772834, |
| "mean_token_accuracy_utility": 0.7039664490029299, |
| "step": 2085 |
| }, |
| { |
| "epoch": 0.7732149463559008, |
| "grad_norm": 1.6484375, |
| "learning_rate": 2.9747504446269447e-06, |
| "loss": 1.8781, |
| "mean_token_accuracy": 0.77014145282265, |
| "mean_token_accuracy_utility": 0.7081794413009034, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.7750647428782834, |
| "grad_norm": 1.6796875, |
| "learning_rate": 2.9289321881345257e-06, |
| "loss": 1.9555, |
| "mean_token_accuracy": 0.7571760716637137, |
| "mean_token_accuracy_utility": 0.7044843682331328, |
| "step": 2095 |
| }, |
| { |
| "epoch": 0.7769145394006659, |
| "grad_norm": 1.7109375, |
| "learning_rate": 2.883408914042435e-06, |
| "loss": 1.9388, |
| "mean_token_accuracy": 0.7496258623781421, |
| "mean_token_accuracy_utility": 0.7107954801732036, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.7769145394006659, |
| "eval_mean_token_accuracy": 0.7532181699492665, |
| "eval_reasoning_loss": 0.8301348090171814, |
| "eval_reasoning_runtime": 29.3176, |
| "eval_reasoning_samples_per_second": 4.4, |
| "eval_reasoning_steps_per_second": 1.126, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.7769145394006659, |
| "eval_mean_token_accuracy": 0.7011768568564241, |
| "eval_utility_loss": 1.1546157598495483, |
| "eval_utility_runtime": 11.0748, |
| "eval_utility_samples_per_second": 4.424, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.7787643359230485, |
| "grad_norm": 1.765625, |
| "learning_rate": 2.838182521436498e-06, |
| "loss": 1.9364, |
| "mean_token_accuracy": 0.7620284496842102, |
| "mean_token_accuracy_utility": 0.7004424004595196, |
| "step": 2105 |
| }, |
| { |
| "epoch": 0.780614132445431, |
| "grad_norm": 1.6875, |
| "learning_rate": 2.7932548970175977e-06, |
| "loss": 1.8934, |
| "mean_token_accuracy": 0.7526547136061004, |
| "mean_token_accuracy_utility": 0.7207447186097956, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.7824639289678136, |
| "grad_norm": 1.6953125, |
| "learning_rate": 2.74862791502295e-06, |
| "loss": 1.8695, |
| "mean_token_accuracy": 0.7680834455522578, |
| "mean_token_accuracy_utility": 0.7139362727375528, |
| "step": 2115 |
| }, |
| { |
| "epoch": 0.7843137254901961, |
| "grad_norm": 1.8671875, |
| "learning_rate": 2.7043034371479404e-06, |
| "loss": 1.8894, |
| "mean_token_accuracy": 0.7548679268583969, |
| "mean_token_accuracy_utility": 0.7173106831589436, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.7861635220125787, |
| "grad_norm": 1.6796875, |
| "learning_rate": 2.660283312468438e-06, |
| "loss": 1.8803, |
| "mean_token_accuracy": 0.7786032120128247, |
| "mean_token_accuracy_utility": 0.7038037787431913, |
| "step": 2125 |
| }, |
| { |
| "epoch": 0.7880133185349611, |
| "grad_norm": 1.734375, |
| "learning_rate": 2.616569377363668e-06, |
| "loss": 2.039, |
| "mean_token_accuracy": 0.7505060410465748, |
| "mean_token_accuracy_utility": 0.6934607431734491, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.7898631150573437, |
| "grad_norm": 1.671875, |
| "learning_rate": 2.573163455439601e-06, |
| "loss": 1.9254, |
| "mean_token_accuracy": 0.761692254609709, |
| "mean_token_accuracy_utility": 0.7112574677528481, |
| "step": 2135 |
| }, |
| { |
| "epoch": 0.7917129115797262, |
| "grad_norm": 1.8046875, |
| "learning_rate": 2.530067357452892e-06, |
| "loss": 2.0186, |
| "mean_token_accuracy": 0.7616279770372245, |
| "mean_token_accuracy_utility": 0.6918121335784461, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.7935627081021087, |
| "grad_norm": 1.640625, |
| "learning_rate": 2.4872828812353146e-06, |
| "loss": 1.7576, |
| "mean_token_accuracy": 0.7683284938353273, |
| "mean_token_accuracy_utility": 0.7334629680685696, |
| "step": 2145 |
| }, |
| { |
| "epoch": 0.7954125046244913, |
| "grad_norm": 1.7109375, |
| "learning_rate": 2.4448118116187847e-06, |
| "loss": 1.9327, |
| "mean_token_accuracy": 0.7513853014666679, |
| "mean_token_accuracy_utility": 0.7079894908210524, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.7972623011468738, |
| "grad_norm": 1.9453125, |
| "learning_rate": 2.402655920360889e-06, |
| "loss": 1.8946, |
| "mean_token_accuracy": 0.7654736697278233, |
| "mean_token_accuracy_utility": 0.7090624634562489, |
| "step": 2155 |
| }, |
| { |
| "epoch": 0.7991120976692564, |
| "grad_norm": 1.8046875, |
| "learning_rate": 2.360816966070986e-06, |
| "loss": 1.867, |
| "mean_token_accuracy": 0.7667388659761512, |
| "mean_token_accuracy_utility": 0.7141480751386007, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.8009618941916389, |
| "grad_norm": 1.6484375, |
| "learning_rate": 2.3192966941368247e-06, |
| "loss": 1.9212, |
| "mean_token_accuracy": 0.7605006998580272, |
| "mean_token_accuracy_utility": 0.707565870328944, |
| "step": 2165 |
| }, |
| { |
| "epoch": 0.8028116907140215, |
| "grad_norm": 1.6796875, |
| "learning_rate": 2.2780968366517476e-06, |
| "loss": 1.9154, |
| "mean_token_accuracy": 0.7773801587519398, |
| "mean_token_accuracy_utility": 0.6948831839253163, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.804661487236404, |
| "grad_norm": 1.6171875, |
| "learning_rate": 2.237219112342426e-06, |
| "loss": 1.9199, |
| "mean_token_accuracy": 0.7664490678990374, |
| "mean_token_accuracy_utility": 0.7029069910422057, |
| "step": 2175 |
| }, |
| { |
| "epoch": 0.8065112837587866, |
| "grad_norm": 1.7578125, |
| "learning_rate": 2.196665226497169e-06, |
| "loss": 1.9747, |
| "mean_token_accuracy": 0.7596191886661391, |
| "mean_token_accuracy_utility": 0.6961392234732143, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.808361080281169, |
| "grad_norm": 1.75, |
| "learning_rate": 2.156436870894767e-06, |
| "loss": 1.9949, |
| "mean_token_accuracy": 0.7523627290250667, |
| "mean_token_accuracy_utility": 0.6978805242636326, |
| "step": 2185 |
| }, |
| { |
| "epoch": 0.8102108768035516, |
| "grad_norm": 1.609375, |
| "learning_rate": 2.116535723733938e-06, |
| "loss": 1.8348, |
| "mean_token_accuracy": 0.7721080082256572, |
| "mean_token_accuracy_utility": 0.7101181129511384, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.8120606733259341, |
| "grad_norm": 1.875, |
| "learning_rate": 2.0769634495632986e-06, |
| "loss": 1.9795, |
| "mean_token_accuracy": 0.7618228153249338, |
| "mean_token_accuracy_utility": 0.6962897334346339, |
| "step": 2195 |
| }, |
| { |
| "epoch": 0.8139104698483167, |
| "grad_norm": 1.890625, |
| "learning_rate": 2.037721699211933e-06, |
| "loss": 1.9564, |
| "mean_token_accuracy": 0.7534863688483261, |
| "mean_token_accuracy_utility": 0.7096826361700221, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.8139104698483167, |
| "eval_mean_token_accuracy": 0.7532181751559996, |
| "eval_reasoning_loss": 0.8299828767776489, |
| "eval_reasoning_runtime": 29.3466, |
| "eval_reasoning_samples_per_second": 4.396, |
| "eval_reasoning_steps_per_second": 1.124, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.8139104698483167, |
| "eval_mean_token_accuracy": 0.7012964187485011, |
| "eval_utility_loss": 1.1546502113342285, |
| "eval_utility_runtime": 11.0916, |
| "eval_utility_samples_per_second": 4.418, |
| "eval_utility_steps_per_second": 1.172, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.8157602663706992, |
| "grad_norm": 1.609375, |
| "learning_rate": 1.998812109720535e-06, |
| "loss": 1.892, |
| "mean_token_accuracy": 0.7575336869803052, |
| "mean_token_accuracy_utility": 0.7133229299717206, |
| "step": 2205 |
| }, |
| { |
| "epoch": 0.8176100628930818, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.9602363042730964e-06, |
| "loss": 1.944, |
| "mean_token_accuracy": 0.7504439500376453, |
| "mean_token_accuracy_utility": 0.7114402930933783, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.8194598594154643, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.921995892129208e-06, |
| "loss": 1.8249, |
| "mean_token_accuracy": 0.7634975834939263, |
| "mean_token_accuracy_utility": 0.725392573927816, |
| "step": 2215 |
| }, |
| { |
| "epoch": 0.8213096559378469, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.8840924685569162e-06, |
| "loss": 1.8801, |
| "mean_token_accuracy": 0.7684288663164015, |
| "mean_token_accuracy_utility": 0.7096136823452641, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.8231594524602294, |
| "grad_norm": 1.78125, |
| "learning_rate": 1.8465276147661905e-06, |
| "loss": 1.8677, |
| "mean_token_accuracy": 0.776187547406066, |
| "mean_token_accuracy_utility": 0.705145679061028, |
| "step": 2225 |
| }, |
| { |
| "epoch": 0.825009248982612, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.809302897842934e-06, |
| "loss": 1.927, |
| "mean_token_accuracy": 0.7602545554810743, |
| "mean_token_accuracy_utility": 0.7077159229055218, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.8268590455049944, |
| "grad_norm": 1.734375, |
| "learning_rate": 1.7724198706836372e-06, |
| "loss": 1.8592, |
| "mean_token_accuracy": 0.7736779814200669, |
| "mean_token_accuracy_utility": 0.7143747357105481, |
| "step": 2235 |
| }, |
| { |
| "epoch": 0.8287088420273769, |
| "grad_norm": 1.7734375, |
| "learning_rate": 1.7358800719305735e-06, |
| "loss": 1.9124, |
| "mean_token_accuracy": 0.7557291282219468, |
| "mean_token_accuracy_utility": 0.7100780677980001, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.8305586385497595, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.6996850259076303e-06, |
| "loss": 1.9758, |
| "mean_token_accuracy": 0.7677954026172016, |
| "mean_token_accuracy_utility": 0.6939027622435243, |
| "step": 2245 |
| }, |
| { |
| "epoch": 0.832408435072142, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.6638362425567012e-06, |
| "loss": 2.0137, |
| "mean_token_accuracy": 0.7509628112058467, |
| "mean_token_accuracy_utility": 0.6921672781412253, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.8342582315945246, |
| "grad_norm": 1.59375, |
| "learning_rate": 1.6283352173747148e-06, |
| "loss": 1.8811, |
| "mean_token_accuracy": 0.7643524335583244, |
| "mean_token_accuracy_utility": 0.713947201805369, |
| "step": 2255 |
| }, |
| { |
| "epoch": 0.8361080281169071, |
| "grad_norm": 1.71875, |
| "learning_rate": 1.5931834313512307e-06, |
| "loss": 1.9772, |
| "mean_token_accuracy": 0.7557003153021895, |
| "mean_token_accuracy_utility": 0.6996604197342927, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.8379578246392897, |
| "grad_norm": 1.6640625, |
| "learning_rate": 1.5583823509066665e-06, |
| "loss": 1.8774, |
| "mean_token_accuracy": 0.7619420369665175, |
| "mean_token_accuracy_utility": 0.7170395183079501, |
| "step": 2265 |
| }, |
| { |
| "epoch": 0.8398076211616722, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.5239334278311214e-06, |
| "loss": 1.862, |
| "mean_token_accuracy": 0.7647236985673922, |
| "mean_token_accuracy_utility": 0.7183922205059409, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.8416574176840548, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.489838099223816e-06, |
| "loss": 1.8894, |
| "mean_token_accuracy": 0.7638369910482289, |
| "mean_token_accuracy_utility": 0.7108152518546073, |
| "step": 2275 |
| }, |
| { |
| "epoch": 0.8435072142064373, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.4560977874331284e-06, |
| "loss": 1.8603, |
| "mean_token_accuracy": 0.7690627342338291, |
| "mean_token_accuracy_utility": 0.7143770672794747, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.8453570107288199, |
| "grad_norm": 1.7109375, |
| "learning_rate": 1.4227138999972801e-06, |
| "loss": 1.88, |
| "mean_token_accuracy": 0.7696150039797313, |
| "mean_token_accuracy_utility": 0.7117503830040024, |
| "step": 2285 |
| }, |
| { |
| "epoch": 0.8472068072512023, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.389687829585592e-06, |
| "loss": 1.9645, |
| "mean_token_accuracy": 0.7549603976198982, |
| "mean_token_accuracy_utility": 0.7078659440200237, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.8490566037735849, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.3570209539404067e-06, |
| "loss": 1.9144, |
| "mean_token_accuracy": 0.7727907756843401, |
| "mean_token_accuracy_utility": 0.6988409395156993, |
| "step": 2295 |
| }, |
| { |
| "epoch": 0.8509064002959674, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.3247146358196073e-06, |
| "loss": 1.9445, |
| "mean_token_accuracy": 0.7604372739462788, |
| "mean_token_accuracy_utility": 0.703307871492025, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.8509064002959674, |
| "eval_mean_token_accuracy": 0.7531160386886891, |
| "eval_reasoning_loss": 0.8299267292022705, |
| "eval_reasoning_runtime": 29.3298, |
| "eval_reasoning_samples_per_second": 4.398, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.8509064002959674, |
| "eval_mean_token_accuracy": 0.7012193655708163, |
| "eval_utility_loss": 1.154624581336975, |
| "eval_utility_runtime": 11.0671, |
| "eval_utility_samples_per_second": 4.428, |
| "eval_utility_steps_per_second": 1.175, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.85275619681835, |
| "grad_norm": 1.703125, |
| "learning_rate": 1.2927702229397633e-06, |
| "loss": 1.8945, |
| "mean_token_accuracy": 0.7577477420479618, |
| "mean_token_accuracy_utility": 0.7131463571382894, |
| "step": 2305 |
| }, |
| { |
| "epoch": 0.8546059933407325, |
| "grad_norm": 1.625, |
| "learning_rate": 1.261189047919913e-06, |
| "loss": 1.8959, |
| "mean_token_accuracy": 0.7587735678487302, |
| "mean_token_accuracy_utility": 0.712902757915149, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.8564557898631151, |
| "grad_norm": 1.640625, |
| "learning_rate": 1.2299724282259685e-06, |
| "loss": 1.9343, |
| "mean_token_accuracy": 0.7574266863266272, |
| "mean_token_accuracy_utility": 0.7017872266654945, |
| "step": 2315 |
| }, |
| { |
| "epoch": 0.8583055863854976, |
| "grad_norm": 2.046875, |
| "learning_rate": 1.1991216661157611e-06, |
| "loss": 1.9325, |
| "mean_token_accuracy": 0.7765085398920788, |
| "mean_token_accuracy_utility": 0.6955550655460147, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.8601553829078802, |
| "grad_norm": 1.7265625, |
| "learning_rate": 1.1686380485847027e-06, |
| "loss": 1.8967, |
| "mean_token_accuracy": 0.7567331533851641, |
| "mean_token_accuracy_utility": 0.7142990182682654, |
| "step": 2325 |
| }, |
| { |
| "epoch": 0.8620051794302627, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.1385228473121113e-06, |
| "loss": 1.8613, |
| "mean_token_accuracy": 0.7753290355363879, |
| "mean_token_accuracy_utility": 0.7089441752678871, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.8638549759526452, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.1087773186081474e-06, |
| "loss": 1.9476, |
| "mean_token_accuracy": 0.7531926600517935, |
| "mean_token_accuracy_utility": 0.7082835067589139, |
| "step": 2335 |
| }, |
| { |
| "epoch": 0.8657047724750278, |
| "grad_norm": 1.5859375, |
| "learning_rate": 1.0794027033614175e-06, |
| "loss": 1.8806, |
| "mean_token_accuracy": 0.7654484462902571, |
| "mean_token_accuracy_utility": 0.7138897225530274, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.8675545689974102, |
| "grad_norm": 1.84375, |
| "learning_rate": 1.0504002269871927e-06, |
| "loss": 1.8973, |
| "mean_token_accuracy": 0.7699482805639639, |
| "mean_token_accuracy_utility": 0.7053482200422982, |
| "step": 2345 |
| }, |
| { |
| "epoch": 0.8694043655197928, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.0217710993763086e-06, |
| "loss": 1.8263, |
| "mean_token_accuracy": 0.7784550760097976, |
| "mean_token_accuracy_utility": 0.7178813805127147, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.8712541620421753, |
| "grad_norm": 1.6484375, |
| "learning_rate": 9.935165148446658e-07, |
| "loss": 1.8999, |
| "mean_token_accuracy": 0.7500060322879973, |
| "mean_token_accuracy_utility": 0.7182965800976913, |
| "step": 2355 |
| }, |
| { |
| "epoch": 0.8731039585645579, |
| "grad_norm": 1.6875, |
| "learning_rate": 9.656376520834365e-07, |
| "loss": 1.9661, |
| "mean_token_accuracy": 0.7619279301321624, |
| "mean_token_accuracy_utility": 0.699871445479563, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.8749537550869404, |
| "grad_norm": 1.6875, |
| "learning_rate": 9.381356741098702e-07, |
| "loss": 1.8381, |
| "mean_token_accuracy": 0.7773417956277906, |
| "mean_token_accuracy_utility": 0.7119904199755409, |
| "step": 2365 |
| }, |
| { |
| "epoch": 0.876803551609323, |
| "grad_norm": 1.7421875, |
| "learning_rate": 9.110117282187892e-07, |
| "loss": 1.8476, |
| "mean_token_accuracy": 0.7833789407563538, |
| "mean_token_accuracy_utility": 0.7065688353768526, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.8786533481317055, |
| "grad_norm": 1.6328125, |
| "learning_rate": 8.842669459347186e-07, |
| "loss": 1.8307, |
| "mean_token_accuracy": 0.759646509327443, |
| "mean_token_accuracy_utility": 0.723392834694595, |
| "step": 2375 |
| }, |
| { |
| "epoch": 0.8805031446540881, |
| "grad_norm": 1.6484375, |
| "learning_rate": 8.579024429646932e-07, |
| "loss": 1.9283, |
| "mean_token_accuracy": 0.7562262974628606, |
| "mean_token_accuracy_utility": 0.70509767797638, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.8823529411764706, |
| "grad_norm": 1.5859375, |
| "learning_rate": 8.319193191517016e-07, |
| "loss": 1.878, |
| "mean_token_accuracy": 0.7652058099089664, |
| "mean_token_accuracy_utility": 0.713667299897729, |
| "step": 2385 |
| }, |
| { |
| "epoch": 0.8842027376988532, |
| "grad_norm": 1.6796875, |
| "learning_rate": 8.063186584288108e-07, |
| "loss": 1.9706, |
| "mean_token_accuracy": 0.7659182674641614, |
| "mean_token_accuracy_utility": 0.696882584776336, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.8860525342212356, |
| "grad_norm": 1.6796875, |
| "learning_rate": 7.81101528773951e-07, |
| "loss": 1.9181, |
| "mean_token_accuracy": 0.7557278168115091, |
| "mean_token_accuracy_utility": 0.7099611906820742, |
| "step": 2395 |
| }, |
| { |
| "epoch": 0.8879023307436182, |
| "grad_norm": 1.71875, |
| "learning_rate": 7.562689821653568e-07, |
| "loss": 1.901, |
| "mean_token_accuracy": 0.7705982155110556, |
| "mean_token_accuracy_utility": 0.7037234895126524, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.8879023307436182, |
| "eval_mean_token_accuracy": 0.7533720855201408, |
| "eval_reasoning_loss": 0.8300442099571228, |
| "eval_reasoning_runtime": 29.3163, |
| "eval_reasoning_samples_per_second": 4.4, |
| "eval_reasoning_steps_per_second": 1.126, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.8879023307436182, |
| "eval_mean_token_accuracy": 0.7013070771208915, |
| "eval_utility_loss": 1.1546061038970947, |
| "eval_utility_runtime": 11.0604, |
| "eval_utility_samples_per_second": 4.43, |
| "eval_utility_steps_per_second": 1.175, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.8897521272660007, |
| "grad_norm": 1.7265625, |
| "learning_rate": 7.318220545376842e-07, |
| "loss": 1.9127, |
| "mean_token_accuracy": 0.76946916046836, |
| "mean_token_accuracy_utility": 0.7027661967055051, |
| "step": 2405 |
| }, |
| { |
| "epoch": 0.8916019237883833, |
| "grad_norm": 1.6640625, |
| "learning_rate": 7.077617657387936e-07, |
| "loss": 1.8914, |
| "mean_token_accuracy": 0.7622229913728239, |
| "mean_token_accuracy_utility": 0.7128095803065368, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.8934517203107658, |
| "grad_norm": 1.7421875, |
| "learning_rate": 6.840891194872112e-07, |
| "loss": 1.996, |
| "mean_token_accuracy": 0.7574201159733877, |
| "mean_token_accuracy_utility": 0.6970374273467271, |
| "step": 2415 |
| }, |
| { |
| "epoch": 0.8953015168331484, |
| "grad_norm": 1.671875, |
| "learning_rate": 6.608051033302476e-07, |
| "loss": 1.8833, |
| "mean_token_accuracy": 0.7590981549860195, |
| "mean_token_accuracy_utility": 0.7108745549232587, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.8971513133555309, |
| "grad_norm": 1.6875, |
| "learning_rate": 6.379106886028086e-07, |
| "loss": 1.8269, |
| "mean_token_accuracy": 0.7769068706934349, |
| "mean_token_accuracy_utility": 0.716204912399294, |
| "step": 2425 |
| }, |
| { |
| "epoch": 0.8990011098779135, |
| "grad_norm": 1.7578125, |
| "learning_rate": 6.154068303868666e-07, |
| "loss": 1.889, |
| "mean_token_accuracy": 0.7657627822691999, |
| "mean_token_accuracy_utility": 0.7086640581576024, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.900850906400296, |
| "grad_norm": 1.671875, |
| "learning_rate": 5.932944674716279e-07, |
| "loss": 1.9149, |
| "mean_token_accuracy": 0.7634639544773816, |
| "mean_token_accuracy_utility": 0.7030951209106451, |
| "step": 2435 |
| }, |
| { |
| "epoch": 0.9027007029226785, |
| "grad_norm": 1.7578125, |
| "learning_rate": 5.715745223143566e-07, |
| "loss": 1.9645, |
| "mean_token_accuracy": 0.7600048934570696, |
| "mean_token_accuracy_utility": 0.6976821210901443, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.904550499445061, |
| "grad_norm": 1.6015625, |
| "learning_rate": 5.502479010019046e-07, |
| "loss": 1.7964, |
| "mean_token_accuracy": 0.7746203264077807, |
| "mean_token_accuracy_utility": 0.716293509962282, |
| "step": 2445 |
| }, |
| { |
| "epoch": 0.9064002959674435, |
| "grad_norm": 1.7109375, |
| "learning_rate": 5.293154932129019e-07, |
| "loss": 2.0267, |
| "mean_token_accuracy": 0.7606274234688003, |
| "mean_token_accuracy_utility": 0.6900432800916387, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.9082500924898261, |
| "grad_norm": 1.7265625, |
| "learning_rate": 5.087781721806539e-07, |
| "loss": 1.8828, |
| "mean_token_accuracy": 0.7755113914311129, |
| "mean_token_accuracy_utility": 0.7059450691135949, |
| "step": 2455 |
| }, |
| { |
| "epoch": 0.9100998890122086, |
| "grad_norm": 1.7109375, |
| "learning_rate": 4.886367946567005e-07, |
| "loss": 1.9695, |
| "mean_token_accuracy": 0.757596735530434, |
| "mean_token_accuracy_utility": 0.7007142502518043, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.9119496855345912, |
| "grad_norm": 1.78125, |
| "learning_rate": 4.6889220087508514e-07, |
| "loss": 1.9029, |
| "mean_token_accuracy": 0.7750377322476185, |
| "mean_token_accuracy_utility": 0.7031355867231781, |
| "step": 2465 |
| }, |
| { |
| "epoch": 0.9137994820569737, |
| "grad_norm": 1.7265625, |
| "learning_rate": 4.4954521451729626e-07, |
| "loss": 1.8042, |
| "mean_token_accuracy": 0.790967029093871, |
| "mean_token_accuracy_utility": 0.7074859497444088, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.9156492785793563, |
| "grad_norm": 1.6640625, |
| "learning_rate": 4.305966426779118e-07, |
| "loss": 1.863, |
| "mean_token_accuracy": 0.7625225628164098, |
| "mean_token_accuracy_utility": 0.7229484946105929, |
| "step": 2475 |
| }, |
| { |
| "epoch": 0.9174990751017388, |
| "grad_norm": 1.71875, |
| "learning_rate": 4.120472758309213e-07, |
| "loss": 1.9012, |
| "mean_token_accuracy": 0.7780731116125541, |
| "mean_token_accuracy_utility": 0.7074578179223155, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.9193488716241214, |
| "grad_norm": 1.6875, |
| "learning_rate": 3.9389788779675806e-07, |
| "loss": 1.9633, |
| "mean_token_accuracy": 0.7537457654946361, |
| "mean_token_accuracy_utility": 0.7053433772241917, |
| "step": 2485 |
| }, |
| { |
| "epoch": 0.9211986681465039, |
| "grad_norm": 1.828125, |
| "learning_rate": 3.761492357100172e-07, |
| "loss": 1.954, |
| "mean_token_accuracy": 0.7613614208065272, |
| "mean_token_accuracy_utility": 0.7001523119154324, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.9230484646688865, |
| "grad_norm": 1.6484375, |
| "learning_rate": 3.588020599878639e-07, |
| "loss": 1.9992, |
| "mean_token_accuracy": 0.7607074234684895, |
| "mean_token_accuracy_utility": 0.6943452792928757, |
| "step": 2495 |
| }, |
| { |
| "epoch": 0.9248982611912689, |
| "grad_norm": 1.7578125, |
| "learning_rate": 3.41857084299152e-07, |
| "loss": 1.8443, |
| "mean_token_accuracy": 0.7725551373073397, |
| "mean_token_accuracy_utility": 0.71925280670418, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.9248982611912689, |
| "eval_mean_token_accuracy": 0.7534294693336302, |
| "eval_reasoning_loss": 0.8300321102142334, |
| "eval_reasoning_runtime": 29.3284, |
| "eval_reasoning_samples_per_second": 4.398, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.9248982611912689, |
| "eval_mean_token_accuracy": 0.7013970369223613, |
| "eval_utility_loss": 1.1545491218566895, |
| "eval_utility_runtime": 11.0901, |
| "eval_utility_samples_per_second": 4.418, |
| "eval_utility_steps_per_second": 1.172, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.9267480577136515, |
| "grad_norm": 1.703125, |
| "learning_rate": 3.2531501553422884e-07, |
| "loss": 1.8963, |
| "mean_token_accuracy": 0.7723618966904187, |
| "mean_token_accuracy_utility": 0.7000879842636227, |
| "step": 2505 |
| }, |
| { |
| "epoch": 0.928597854236034, |
| "grad_norm": 1.8203125, |
| "learning_rate": 3.0917654377545567e-07, |
| "loss": 1.9324, |
| "mean_token_accuracy": 0.7706404315723949, |
| "mean_token_accuracy_utility": 0.6973090499209622, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.9304476507584166, |
| "grad_norm": 1.703125, |
| "learning_rate": 2.9344234226840964e-07, |
| "loss": 1.9376, |
| "mean_token_accuracy": 0.76105847827583, |
| "mean_token_accuracy_utility": 0.7054121315247243, |
| "step": 2515 |
| }, |
| { |
| "epoch": 0.9322974472807991, |
| "grad_norm": 1.78125, |
| "learning_rate": 2.781130673938037e-07, |
| "loss": 1.9508, |
| "mean_token_accuracy": 0.7637577676005167, |
| "mean_token_accuracy_utility": 0.6988380737178091, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.9341472438031817, |
| "grad_norm": 1.7890625, |
| "learning_rate": 2.6318935864010133e-07, |
| "loss": 1.9083, |
| "mean_token_accuracy": 0.7605887695277631, |
| "mean_token_accuracy_utility": 0.7081206644601539, |
| "step": 2525 |
| }, |
| { |
| "epoch": 0.9359970403255642, |
| "grad_norm": 1.6171875, |
| "learning_rate": 2.4867183857684385e-07, |
| "loss": 1.8405, |
| "mean_token_accuracy": 0.7722669253591803, |
| "mean_token_accuracy_utility": 0.7138795962861496, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.9378468368479467, |
| "grad_norm": 1.8046875, |
| "learning_rate": 2.3456111282867178e-07, |
| "loss": 1.953, |
| "mean_token_accuracy": 0.7670061175692017, |
| "mean_token_accuracy_utility": 0.69500238489083, |
| "step": 2535 |
| }, |
| { |
| "epoch": 0.9396966333703293, |
| "grad_norm": 1.6484375, |
| "learning_rate": 2.2085777005006692e-07, |
| "loss": 1.879, |
| "mean_token_accuracy": 0.7677980910484122, |
| "mean_token_accuracy_utility": 0.7159645171707567, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.9415464298927118, |
| "grad_norm": 1.8125, |
| "learning_rate": 2.0756238190078991e-07, |
| "loss": 1.8482, |
| "mean_token_accuracy": 0.7686032444763766, |
| "mean_token_accuracy_utility": 0.7134517743859778, |
| "step": 2545 |
| }, |
| { |
| "epoch": 0.9433962264150944, |
| "grad_norm": 1.9453125, |
| "learning_rate": 1.946755030220393e-07, |
| "loss": 1.9832, |
| "mean_token_accuracy": 0.7490072983413629, |
| "mean_token_accuracy_utility": 0.7049523505044996, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.9452460229374768, |
| "grad_norm": 1.6875, |
| "learning_rate": 1.8219767101330442e-07, |
| "loss": 1.944, |
| "mean_token_accuracy": 0.7684967053732852, |
| "mean_token_accuracy_utility": 0.6977903705987989, |
| "step": 2555 |
| }, |
| { |
| "epoch": 0.9470958194598594, |
| "grad_norm": 1.8671875, |
| "learning_rate": 1.7012940640995012e-07, |
| "loss": 1.8849, |
| "mean_token_accuracy": 0.7628139156548163, |
| "mean_token_accuracy_utility": 0.7120314817716029, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.9489456159822419, |
| "grad_norm": 1.6953125, |
| "learning_rate": 1.5847121266148847e-07, |
| "loss": 1.8562, |
| "mean_token_accuracy": 0.7617193020783564, |
| "mean_token_accuracy_utility": 0.7162649275350278, |
| "step": 2565 |
| }, |
| { |
| "epoch": 0.9507954125046245, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.472235761105878e-07, |
| "loss": 1.8932, |
| "mean_token_accuracy": 0.7826583482318876, |
| "mean_token_accuracy_utility": 0.6964653965582813, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.952645209027007, |
| "grad_norm": 1.8125, |
| "learning_rate": 1.3638696597277678e-07, |
| "loss": 1.9382, |
| "mean_token_accuracy": 0.7605572848541231, |
| "mean_token_accuracy_utility": 0.7038059046095655, |
| "step": 2575 |
| }, |
| { |
| "epoch": 0.9544950055493896, |
| "grad_norm": 1.75, |
| "learning_rate": 1.2596183431687336e-07, |
| "loss": 1.9645, |
| "mean_token_accuracy": 0.7658640055908315, |
| "mean_token_accuracy_utility": 0.6992427416063306, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.9563448020717721, |
| "grad_norm": 1.671875, |
| "learning_rate": 1.159486160461265e-07, |
| "loss": 1.8823, |
| "mean_token_accuracy": 0.7611909852648259, |
| "mean_token_accuracy_utility": 0.7148172899921141, |
| "step": 2585 |
| }, |
| { |
| "epoch": 0.9581945985941547, |
| "grad_norm": 1.8046875, |
| "learning_rate": 1.0634772888007073e-07, |
| "loss": 1.9221, |
| "mean_token_accuracy": 0.7688862889261017, |
| "mean_token_accuracy_utility": 0.7004788401267631, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.9600443951165372, |
| "grad_norm": 1.6796875, |
| "learning_rate": 9.715957333710447e-08, |
| "loss": 1.7664, |
| "mean_token_accuracy": 0.7717100829011501, |
| "mean_token_accuracy_utility": 0.7275011368738776, |
| "step": 2595 |
| }, |
| { |
| "epoch": 0.9618941916389198, |
| "grad_norm": 1.6875, |
| "learning_rate": 8.838453271777347e-08, |
| "loss": 1.9183, |
| "mean_token_accuracy": 0.7752028138790468, |
| "mean_token_accuracy_utility": 0.6944930301403665, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.9618941916389198, |
| "eval_mean_token_accuracy": 0.753177723287291, |
| "eval_reasoning_loss": 0.8300263285636902, |
| "eval_reasoning_runtime": 29.3265, |
| "eval_reasoning_samples_per_second": 4.399, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.9618941916389198, |
| "eval_mean_token_accuracy": 0.7012780911514778, |
| "eval_utility_loss": 1.15445876121521, |
| "eval_utility_runtime": 11.0811, |
| "eval_utility_samples_per_second": 4.422, |
| "eval_utility_steps_per_second": 1.173, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.9637439881613022, |
| "grad_norm": 1.734375, |
| "learning_rate": 8.002297308879359e-08, |
| "loss": 1.8879, |
| "mean_token_accuracy": 0.7558604952012382, |
| "mean_token_accuracy_utility": 0.7169823439200368, |
| "step": 2605 |
| }, |
| { |
| "epoch": 0.9655937846836848, |
| "grad_norm": 1.65625, |
| "learning_rate": 7.207524326776628e-08, |
| "loss": 1.9021, |
| "mean_token_accuracy": 0.7718762514859595, |
| "mean_token_accuracy_utility": 0.6995048127381205, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.9674435812060673, |
| "grad_norm": 1.671875, |
| "learning_rate": 6.454167480863694e-08, |
| "loss": 1.9737, |
| "mean_token_accuracy": 0.7536406401407107, |
| "mean_token_accuracy_utility": 0.7010452116455725, |
| "step": 2615 |
| }, |
| { |
| "epoch": 0.9692933777284499, |
| "grad_norm": 1.671875, |
| "learning_rate": 5.7422581987856e-08, |
| "loss": 1.9249, |
| "mean_token_accuracy": 0.7614680395118091, |
| "mean_token_accuracy_utility": 0.7038131554658986, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.9711431742508324, |
| "grad_norm": 1.625, |
| "learning_rate": 5.0718261791274924e-08, |
| "loss": 1.9717, |
| "mean_token_accuracy": 0.7495392927026996, |
| "mean_token_accuracy_utility": 0.704763374617239, |
| "step": 2625 |
| }, |
| { |
| "epoch": 0.9729929707732149, |
| "grad_norm": 1.7265625, |
| "learning_rate": 4.4428993901753924e-08, |
| "loss": 1.8178, |
| "mean_token_accuracy": 0.769242972315282, |
| "mean_token_accuracy_utility": 0.7219520804767903, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.9748427672955975, |
| "grad_norm": 1.75, |
| "learning_rate": 3.8555040687493494e-08, |
| "loss": 1.8668, |
| "mean_token_accuracy": 0.771706065448497, |
| "mean_token_accuracy_utility": 0.7102138968291358, |
| "step": 2635 |
| }, |
| { |
| "epoch": 0.97669256381798, |
| "grad_norm": 1.671875, |
| "learning_rate": 3.309664719108874e-08, |
| "loss": 1.9215, |
| "mean_token_accuracy": 0.7791101118954231, |
| "mean_token_accuracy_utility": 0.6962919595070947, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.9785423603403626, |
| "grad_norm": 1.7890625, |
| "learning_rate": 2.805404111931198e-08, |
| "loss": 1.9227, |
| "mean_token_accuracy": 0.7658102511811766, |
| "mean_token_accuracy_utility": 0.7028243093619838, |
| "step": 2645 |
| }, |
| { |
| "epoch": 0.9803921568627451, |
| "grad_norm": 1.6953125, |
| "learning_rate": 2.3427432833607046e-08, |
| "loss": 1.878, |
| "mean_token_accuracy": 0.7645758369515584, |
| "mean_token_accuracy_utility": 0.7104918994511704, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.9822419533851277, |
| "grad_norm": 1.6171875, |
| "learning_rate": 1.9217015341318478e-08, |
| "loss": 1.9501, |
| "mean_token_accuracy": 0.7521906379466751, |
| "mean_token_accuracy_utility": 0.7092303330245882, |
| "step": 2655 |
| }, |
| { |
| "epoch": 0.9840917499075101, |
| "grad_norm": 1.8203125, |
| "learning_rate": 1.542296428763801e-08, |
| "loss": 1.8919, |
| "mean_token_accuracy": 0.7719197279075914, |
| "mean_token_accuracy_utility": 0.6997915488635911, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.9859415464298927, |
| "grad_norm": 1.7578125, |
| "learning_rate": 1.2045437948275952e-08, |
| "loss": 2.02, |
| "mean_token_accuracy": 0.7571515327038465, |
| "mean_token_accuracy_utility": 0.6885369765304264, |
| "step": 2665 |
| }, |
| { |
| "epoch": 0.9877913429522752, |
| "grad_norm": 1.8046875, |
| "learning_rate": 9.084577222860935e-09, |
| "loss": 1.8555, |
| "mean_token_accuracy": 0.7732471226217539, |
| "mean_token_accuracy_utility": 0.7106250285175941, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.9896411394746578, |
| "grad_norm": 1.734375, |
| "learning_rate": 6.540505629061278e-09, |
| "loss": 1.8769, |
| "mean_token_accuracy": 0.7625008803075453, |
| "mean_token_accuracy_utility": 0.7152767747928865, |
| "step": 2675 |
| }, |
| { |
| "epoch": 0.9914909359970403, |
| "grad_norm": 1.7890625, |
| "learning_rate": 4.413329297430213e-09, |
| "loss": 1.9117, |
| "mean_token_accuracy": 0.7657409246436766, |
| "mean_token_accuracy_utility": 0.7050666307214885, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.9933407325194229, |
| "grad_norm": 1.65625, |
| "learning_rate": 2.7031369669816566e-09, |
| "loss": 1.941, |
| "mean_token_accuracy": 0.7462401035629032, |
| "mean_token_accuracy_utility": 0.7152714051431875, |
| "step": 2685 |
| }, |
| { |
| "epoch": 0.9951905290418054, |
| "grad_norm": 1.8359375, |
| "learning_rate": 1.4099999814853883e-09, |
| "loss": 1.9823, |
| "mean_token_accuracy": 0.7626959951066943, |
| "mean_token_accuracy_utility": 0.6958993751675076, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.997040325564188, |
| "grad_norm": 1.671875, |
| "learning_rate": 5.339722864927677e-10, |
| "loss": 1.9301, |
| "mean_token_accuracy": 0.7595753298323694, |
| "mean_token_accuracy_utility": 0.7047228975554161, |
| "step": 2695 |
| }, |
| { |
| "epoch": 0.9988901220865705, |
| "grad_norm": 1.859375, |
| "learning_rate": 7.509042708631065e-11, |
| "loss": 1.976, |
| "mean_token_accuracy": 0.7599278147335686, |
| "mean_token_accuracy_utility": 0.6961616835811532, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.9988901220865705, |
| "eval_mean_token_accuracy": 0.7534720318288755, |
| "eval_reasoning_loss": 0.829900324344635, |
| "eval_reasoning_runtime": 29.3367, |
| "eval_reasoning_samples_per_second": 4.397, |
| "eval_reasoning_steps_per_second": 1.125, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.9988901220865705, |
| "eval_mean_token_accuracy": 0.701419298002488, |
| "eval_utility_loss": 1.154617190361023, |
| "eval_utility_runtime": 11.0722, |
| "eval_utility_samples_per_second": 4.425, |
| "eval_utility_steps_per_second": 1.174, |
| "step": 2700 |
| }, |
| { |
| "epoch": 1.0, |
| "mean_token_accuracy": 0.7603333790417119, |
| "mean_token_accuracy_utility": 0.7230905196106008, |
| "step": 2703, |
| "total_flos": 0.0, |
| "train_loss": 1.963218881252118, |
| "train_runtime": 27333.3295, |
| "train_samples_per_second": 0.791, |
| "train_steps_per_second": 0.099 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 2703, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
