{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.28, "eval_steps": 500, "global_step": 4000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "entropy": 6.984166097640991, "epoch": 0.0016, "grad_norm": 35.75, "learning_rate": 2e-06, "loss": 6.7632, "mean_token_accuracy": 0.0006622516550123692, "num_tokens": 2386.0, "step": 5 }, { "entropy": 6.983382844924927, "epoch": 0.0032, "grad_norm": 35.0, "learning_rate": 4.5e-06, "loss": 6.5137, "mean_token_accuracy": 0.00911798703018576, "num_tokens": 4470.0, "step": 10 }, { "entropy": 6.92952790260315, "epoch": 0.0048, "grad_norm": 20.125, "learning_rate": 7e-06, "loss": 5.9157, "mean_token_accuracy": 0.13627840019762516, "num_tokens": 6676.0, "step": 15 }, { "entropy": 6.516384649276733, "epoch": 0.0064, "grad_norm": 11.0, "learning_rate": 9.5e-06, "loss": 5.2231, "mean_token_accuracy": 0.18535970747470856, "num_tokens": 8950.0, "step": 20 }, { "entropy": 5.922508668899536, "epoch": 0.008, "grad_norm": 6.875, "learning_rate": 1.2e-05, "loss": 4.7583, "mean_token_accuracy": 0.22232983112335206, "num_tokens": 11297.0, "step": 25 }, { "entropy": 5.376971960067749, "epoch": 0.0096, "grad_norm": 6.9375, "learning_rate": 1.4500000000000002e-05, "loss": 4.4331, "mean_token_accuracy": 0.2609219327569008, "num_tokens": 13319.0, "step": 30 }, { "entropy": 5.280996513366699, "epoch": 0.0112, "grad_norm": 6.21875, "learning_rate": 1.7000000000000003e-05, "loss": 4.3976, "mean_token_accuracy": 0.2600315988063812, "num_tokens": 15587.0, "step": 35 }, { "entropy": 5.127728080749511, "epoch": 0.0128, "grad_norm": 9.875, "learning_rate": 1.95e-05, "loss": 4.2436, "mean_token_accuracy": 0.2905985027551651, "num_tokens": 17930.0, "step": 40 }, { "entropy": 4.909821891784668, "epoch": 0.0144, "grad_norm": 8.5625, "learning_rate": 2.2e-05, "loss": 4.2283, "mean_token_accuracy": 0.2786957398056984, "num_tokens": 20298.0, "step": 45 }, { "entropy": 4.7780097961425785, "epoch": 0.016, "grad_norm": 15.0625, "learning_rate": 2.4500000000000003e-05, "loss": 4.0488, "mean_token_accuracy": 0.31810986399650576, "num_tokens": 22451.0, "step": 50 }, { "entropy": 4.597752809524536, "epoch": 0.0176, "grad_norm": 9.625, "learning_rate": 2.7e-05, "loss": 3.9814, "mean_token_accuracy": 0.3224632740020752, "num_tokens": 24577.0, "step": 55 }, { "entropy": 4.443562126159668, "epoch": 0.0192, "grad_norm": 11.0, "learning_rate": 2.95e-05, "loss": 3.9052, "mean_token_accuracy": 0.32513262033462526, "num_tokens": 26636.0, "step": 60 }, { "entropy": 4.546718502044678, "epoch": 0.0208, "grad_norm": 8.625, "learning_rate": 3.2e-05, "loss": 3.8759, "mean_token_accuracy": 0.33354621231555937, "num_tokens": 28845.0, "step": 65 }, { "entropy": 4.427032899856568, "epoch": 0.0224, "grad_norm": 10.3125, "learning_rate": 3.4500000000000005e-05, "loss": 3.8679, "mean_token_accuracy": 0.3256209403276443, "num_tokens": 31061.0, "step": 70 }, { "entropy": 4.440110492706299, "epoch": 0.024, "grad_norm": 6.09375, "learning_rate": 3.7e-05, "loss": 3.8439, "mean_token_accuracy": 0.32602192014455794, "num_tokens": 33235.0, "step": 75 }, { "entropy": 4.499065685272217, "epoch": 0.0256, "grad_norm": 6.625, "learning_rate": 3.95e-05, "loss": 3.8513, "mean_token_accuracy": 0.32822578996419904, "num_tokens": 35597.0, "step": 80 }, { "entropy": 4.301044225692749, "epoch": 0.0272, "grad_norm": 9.4375, "learning_rate": 4.2000000000000004e-05, "loss": 3.6838, "mean_token_accuracy": 0.35025333166122435, "num_tokens": 37679.0, "step": 85 }, { "entropy": 4.261509561538697, "epoch": 0.0288, "grad_norm": 5.59375, "learning_rate": 4.45e-05, "loss": 3.605, "mean_token_accuracy": 0.3899139821529388, "num_tokens": 39714.0, "step": 90 }, { "entropy": 4.252111029624939, "epoch": 0.0304, "grad_norm": 7.125, "learning_rate": 4.7000000000000004e-05, "loss": 3.5804, "mean_token_accuracy": 0.3761704981327057, "num_tokens": 41888.0, "step": 95 }, { "entropy": 4.292388868331909, "epoch": 0.032, "grad_norm": 7.5625, "learning_rate": 4.9500000000000004e-05, "loss": 3.641, "mean_token_accuracy": 0.3718427151441574, "num_tokens": 44202.0, "step": 100 }, { "entropy": 4.046170234680176, "epoch": 0.0336, "grad_norm": 6.25, "learning_rate": 5.2e-05, "loss": 3.5391, "mean_token_accuracy": 0.3826078325510025, "num_tokens": 46269.0, "step": 105 }, { "entropy": 4.0796456098556515, "epoch": 0.0352, "grad_norm": 6.96875, "learning_rate": 5.45e-05, "loss": 3.5599, "mean_token_accuracy": 0.3781130462884903, "num_tokens": 48305.0, "step": 110 }, { "entropy": 4.137033724784851, "epoch": 0.0368, "grad_norm": 5.5625, "learning_rate": 5.7e-05, "loss": 3.4631, "mean_token_accuracy": 0.3850142925977707, "num_tokens": 50511.0, "step": 115 }, { "entropy": 4.137476301193237, "epoch": 0.0384, "grad_norm": 8.0, "learning_rate": 5.9499999999999996e-05, "loss": 3.5881, "mean_token_accuracy": 0.37923752665519717, "num_tokens": 52700.0, "step": 120 }, { "entropy": 4.10472559928894, "epoch": 0.04, "grad_norm": 8.375, "learning_rate": 6.2e-05, "loss": 3.5367, "mean_token_accuracy": 0.3789939880371094, "num_tokens": 54993.0, "step": 125 }, { "entropy": 4.036384010314942, "epoch": 0.0416, "grad_norm": 6.21875, "learning_rate": 6.450000000000001e-05, "loss": 3.4892, "mean_token_accuracy": 0.3903819382190704, "num_tokens": 57188.0, "step": 130 }, { "entropy": 4.002517557144165, "epoch": 0.0432, "grad_norm": 6.65625, "learning_rate": 6.7e-05, "loss": 3.4338, "mean_token_accuracy": 0.3925500184297562, "num_tokens": 59353.0, "step": 135 }, { "entropy": 3.964102292060852, "epoch": 0.0448, "grad_norm": 6.40625, "learning_rate": 6.950000000000001e-05, "loss": 3.3553, "mean_token_accuracy": 0.40215928554534913, "num_tokens": 61474.0, "step": 140 }, { "entropy": 3.9429901838302612, "epoch": 0.0464, "grad_norm": 7.65625, "learning_rate": 7.2e-05, "loss": 3.3097, "mean_token_accuracy": 0.41020109355449674, "num_tokens": 63530.0, "step": 145 }, { "entropy": 3.9518532037734984, "epoch": 0.048, "grad_norm": 4.40625, "learning_rate": 7.45e-05, "loss": 3.4309, "mean_token_accuracy": 0.39309887290000917, "num_tokens": 65670.0, "step": 150 }, { "entropy": 3.8291793346405028, "epoch": 0.0496, "grad_norm": 4.96875, "learning_rate": 7.7e-05, "loss": 3.3886, "mean_token_accuracy": 0.4117222875356674, "num_tokens": 67878.0, "step": 155 }, { "entropy": 3.899980068206787, "epoch": 0.0512, "grad_norm": 8.0, "learning_rate": 7.950000000000001e-05, "loss": 3.3575, "mean_token_accuracy": 0.39695691168308256, "num_tokens": 70104.0, "step": 160 }, { "entropy": 3.858958888053894, "epoch": 0.0528, "grad_norm": 5.1875, "learning_rate": 8.2e-05, "loss": 3.3046, "mean_token_accuracy": 0.40038737654685974, "num_tokens": 72287.0, "step": 165 }, { "entropy": 3.829202103614807, "epoch": 0.0544, "grad_norm": 4.9375, "learning_rate": 8.450000000000001e-05, "loss": 3.3057, "mean_token_accuracy": 0.4086408346891403, "num_tokens": 74630.0, "step": 170 }, { "entropy": 3.813246655464172, "epoch": 0.056, "grad_norm": 5.40625, "learning_rate": 8.7e-05, "loss": 3.3499, "mean_token_accuracy": 0.40505404472351075, "num_tokens": 76837.0, "step": 175 }, { "entropy": 3.950967574119568, "epoch": 0.0576, "grad_norm": 8.625, "learning_rate": 8.95e-05, "loss": 3.3502, "mean_token_accuracy": 0.39242727756500245, "num_tokens": 79091.0, "step": 180 }, { "entropy": 3.72328360080719, "epoch": 0.0592, "grad_norm": 6.125, "learning_rate": 9.2e-05, "loss": 3.2306, "mean_token_accuracy": 0.4187060624361038, "num_tokens": 81255.0, "step": 185 }, { "entropy": 3.7847073793411257, "epoch": 0.0608, "grad_norm": 6.625, "learning_rate": 9.45e-05, "loss": 3.2252, "mean_token_accuracy": 0.40823503136634826, "num_tokens": 83360.0, "step": 190 }, { "entropy": 3.784400296211243, "epoch": 0.0624, "grad_norm": 6.65625, "learning_rate": 9.7e-05, "loss": 3.2711, "mean_token_accuracy": 0.41136759519577026, "num_tokens": 85588.0, "step": 195 }, { "entropy": 3.77215371131897, "epoch": 0.064, "grad_norm": 4.46875, "learning_rate": 9.95e-05, "loss": 3.2348, "mean_token_accuracy": 0.40729652941226957, "num_tokens": 87727.0, "step": 200 }, { "entropy": 3.735856127738953, "epoch": 0.0656, "grad_norm": 4.625, "learning_rate": 0.000102, "loss": 3.2761, "mean_token_accuracy": 0.40915278494358065, "num_tokens": 90191.0, "step": 205 }, { "entropy": 3.5806185722351076, "epoch": 0.0672, "grad_norm": 4.09375, "learning_rate": 0.00010449999999999999, "loss": 3.1817, "mean_token_accuracy": 0.4046794682741165, "num_tokens": 92301.0, "step": 210 }, { "entropy": 3.5795039176940917, "epoch": 0.0688, "grad_norm": 5.625, "learning_rate": 0.000107, "loss": 3.1391, "mean_token_accuracy": 0.412116077542305, "num_tokens": 94397.0, "step": 215 }, { "entropy": 3.703463912010193, "epoch": 0.0704, "grad_norm": 4.75, "learning_rate": 0.0001095, "loss": 3.2693, "mean_token_accuracy": 0.39964946508407595, "num_tokens": 96581.0, "step": 220 }, { "entropy": 3.5505415916442873, "epoch": 0.072, "grad_norm": 4.625, "learning_rate": 0.000112, "loss": 3.1821, "mean_token_accuracy": 0.4107900172472, "num_tokens": 98690.0, "step": 225 }, { "entropy": 3.6934717416763307, "epoch": 0.0736, "grad_norm": 4.0, "learning_rate": 0.0001145, "loss": 3.1325, "mean_token_accuracy": 0.4253169983625412, "num_tokens": 100899.0, "step": 230 }, { "entropy": 3.442733335494995, "epoch": 0.0752, "grad_norm": 2.828125, "learning_rate": 0.00011700000000000001, "loss": 3.2048, "mean_token_accuracy": 0.40409339070320127, "num_tokens": 103262.0, "step": 235 }, { "entropy": 3.621869516372681, "epoch": 0.0768, "grad_norm": 2.578125, "learning_rate": 0.00011949999999999999, "loss": 3.0758, "mean_token_accuracy": 0.433078932762146, "num_tokens": 105354.0, "step": 240 }, { "entropy": 3.5663766860961914, "epoch": 0.0784, "grad_norm": 2.765625, "learning_rate": 0.000122, "loss": 3.1016, "mean_token_accuracy": 0.42279450297355653, "num_tokens": 107499.0, "step": 245 }, { "entropy": 3.3743895292282104, "epoch": 0.08, "grad_norm": 3.859375, "learning_rate": 0.0001245, "loss": 3.1138, "mean_token_accuracy": 0.4299220085144043, "num_tokens": 109611.0, "step": 250 }, { "entropy": 3.6112706661224365, "epoch": 0.0816, "grad_norm": 3.1875, "learning_rate": 0.000127, "loss": 3.1767, "mean_token_accuracy": 0.42330595254898074, "num_tokens": 111816.0, "step": 255 }, { "entropy": 3.48031632900238, "epoch": 0.0832, "grad_norm": 2.65625, "learning_rate": 0.0001295, "loss": 3.0398, "mean_token_accuracy": 0.4351245522499084, "num_tokens": 113994.0, "step": 260 }, { "entropy": 3.4091481447219847, "epoch": 0.0848, "grad_norm": 3.046875, "learning_rate": 0.000132, "loss": 3.0441, "mean_token_accuracy": 0.4238955587148666, "num_tokens": 116191.0, "step": 265 }, { "entropy": 3.3985170125961304, "epoch": 0.0864, "grad_norm": 3.734375, "learning_rate": 0.00013450000000000002, "loss": 3.1052, "mean_token_accuracy": 0.4255074173212051, "num_tokens": 118508.0, "step": 270 }, { "entropy": 3.3496312379837034, "epoch": 0.088, "grad_norm": 3.5, "learning_rate": 0.00013700000000000002, "loss": 3.0903, "mean_token_accuracy": 0.41443932950496676, "num_tokens": 120734.0, "step": 275 }, { "entropy": 3.4435964822769165, "epoch": 0.0896, "grad_norm": 2.796875, "learning_rate": 0.0001395, "loss": 3.1691, "mean_token_accuracy": 0.40538413226604464, "num_tokens": 123071.0, "step": 280 }, { "entropy": 3.4586278915405275, "epoch": 0.0912, "grad_norm": 3.1875, "learning_rate": 0.00014199999999999998, "loss": 3.0854, "mean_token_accuracy": 0.41793314218521116, "num_tokens": 125366.0, "step": 285 }, { "entropy": 3.3301186084747316, "epoch": 0.0928, "grad_norm": 2.625, "learning_rate": 0.0001445, "loss": 2.972, "mean_token_accuracy": 0.4431908309459686, "num_tokens": 127353.0, "step": 290 }, { "entropy": 3.2211655378341675, "epoch": 0.0944, "grad_norm": 2.71875, "learning_rate": 0.000147, "loss": 2.9955, "mean_token_accuracy": 0.43748491406440737, "num_tokens": 129484.0, "step": 295 }, { "entropy": 3.503356146812439, "epoch": 0.096, "grad_norm": 2.96875, "learning_rate": 0.0001495, "loss": 3.0979, "mean_token_accuracy": 0.4168858081102371, "num_tokens": 131727.0, "step": 300 }, { "entropy": 3.267705965042114, "epoch": 0.0976, "grad_norm": 3.515625, "learning_rate": 0.000152, "loss": 3.0941, "mean_token_accuracy": 0.4232072412967682, "num_tokens": 134018.0, "step": 305 }, { "entropy": 3.4302841663360595, "epoch": 0.0992, "grad_norm": 2.34375, "learning_rate": 0.00015450000000000001, "loss": 2.9896, "mean_token_accuracy": 0.43529458045959474, "num_tokens": 136109.0, "step": 310 }, { "entropy": 3.211355781555176, "epoch": 0.1008, "grad_norm": 2.578125, "learning_rate": 0.000157, "loss": 3.037, "mean_token_accuracy": 0.437214794754982, "num_tokens": 138291.0, "step": 315 }, { "entropy": 3.3611859321594237, "epoch": 0.1024, "grad_norm": 2.515625, "learning_rate": 0.0001595, "loss": 3.046, "mean_token_accuracy": 0.42600254714488983, "num_tokens": 140549.0, "step": 320 }, { "entropy": 3.2808714866638184, "epoch": 0.104, "grad_norm": 2.375, "learning_rate": 0.000162, "loss": 3.067, "mean_token_accuracy": 0.41987003684043883, "num_tokens": 142707.0, "step": 325 }, { "entropy": 3.262523865699768, "epoch": 0.1056, "grad_norm": 3.078125, "learning_rate": 0.00016450000000000001, "loss": 3.032, "mean_token_accuracy": 0.4233466386795044, "num_tokens": 144910.0, "step": 330 }, { "entropy": 3.351242733001709, "epoch": 0.1072, "grad_norm": 2.828125, "learning_rate": 0.00016700000000000002, "loss": 3.0574, "mean_token_accuracy": 0.4343830496072769, "num_tokens": 147139.0, "step": 335 }, { "entropy": 3.220554971694946, "epoch": 0.1088, "grad_norm": 2.6875, "learning_rate": 0.00016950000000000003, "loss": 3.1111, "mean_token_accuracy": 0.4186121106147766, "num_tokens": 149489.0, "step": 340 }, { "entropy": 3.2088125944137573, "epoch": 0.1104, "grad_norm": 2.140625, "learning_rate": 0.00017199999999999998, "loss": 3.043, "mean_token_accuracy": 0.41911496222019196, "num_tokens": 151799.0, "step": 345 }, { "entropy": 3.310237908363342, "epoch": 0.112, "grad_norm": 2.109375, "learning_rate": 0.00017449999999999999, "loss": 3.045, "mean_token_accuracy": 0.4296528369188309, "num_tokens": 154108.0, "step": 350 }, { "entropy": 3.082780146598816, "epoch": 0.1136, "grad_norm": 2.96875, "learning_rate": 0.000177, "loss": 3.0274, "mean_token_accuracy": 0.44388274252414706, "num_tokens": 156330.0, "step": 355 }, { "entropy": 3.276983118057251, "epoch": 0.1152, "grad_norm": 2.3125, "learning_rate": 0.0001795, "loss": 3.063, "mean_token_accuracy": 0.42635676860809324, "num_tokens": 158687.0, "step": 360 }, { "entropy": 3.1190796375274656, "epoch": 0.1168, "grad_norm": 2.796875, "learning_rate": 0.000182, "loss": 2.9668, "mean_token_accuracy": 0.437558525800705, "num_tokens": 160746.0, "step": 365 }, { "entropy": 3.095254921913147, "epoch": 0.1184, "grad_norm": 2.171875, "learning_rate": 0.0001845, "loss": 2.9724, "mean_token_accuracy": 0.4385862618684769, "num_tokens": 162867.0, "step": 370 }, { "entropy": 3.2327961921691895, "epoch": 0.12, "grad_norm": 2.15625, "learning_rate": 0.000187, "loss": 3.0294, "mean_token_accuracy": 0.4309984356164932, "num_tokens": 165170.0, "step": 375 }, { "entropy": 3.1149100065231323, "epoch": 0.1216, "grad_norm": 2.28125, "learning_rate": 0.0001895, "loss": 3.0149, "mean_token_accuracy": 0.42866076827049254, "num_tokens": 167445.0, "step": 380 }, { "entropy": 3.082882022857666, "epoch": 0.1232, "grad_norm": 4.28125, "learning_rate": 0.000192, "loss": 3.0066, "mean_token_accuracy": 0.4264851897954941, "num_tokens": 169563.0, "step": 385 }, { "entropy": 3.140112376213074, "epoch": 0.1248, "grad_norm": 3.0625, "learning_rate": 0.0001945, "loss": 3.0293, "mean_token_accuracy": 0.42141772210597994, "num_tokens": 171808.0, "step": 390 }, { "entropy": 3.0665892601013183, "epoch": 0.1264, "grad_norm": 2.953125, "learning_rate": 0.00019700000000000002, "loss": 2.9865, "mean_token_accuracy": 0.436166051030159, "num_tokens": 173924.0, "step": 395 }, { "entropy": 3.121848154067993, "epoch": 0.128, "grad_norm": 3.0625, "learning_rate": 0.00019950000000000002, "loss": 3.0002, "mean_token_accuracy": 0.43142950236797334, "num_tokens": 176225.0, "step": 400 }, { "entropy": 3.202942943572998, "epoch": 0.1296, "grad_norm": 1.5234375, "learning_rate": 0.000202, "loss": 3.0292, "mean_token_accuracy": 0.42796260118484497, "num_tokens": 178624.0, "step": 405 }, { "entropy": 3.020462989807129, "epoch": 0.1312, "grad_norm": 1.8046875, "learning_rate": 0.00020449999999999998, "loss": 2.9524, "mean_token_accuracy": 0.4469933271408081, "num_tokens": 180702.0, "step": 410 }, { "entropy": 3.1144888401031494, "epoch": 0.1328, "grad_norm": 3.328125, "learning_rate": 0.000207, "loss": 3.0272, "mean_token_accuracy": 0.4237643748521805, "num_tokens": 183136.0, "step": 415 }, { "entropy": 3.1383991718292235, "epoch": 0.1344, "grad_norm": 3.453125, "learning_rate": 0.0002095, "loss": 3.0466, "mean_token_accuracy": 0.42285779416561126, "num_tokens": 185284.0, "step": 420 }, { "entropy": 3.106661081314087, "epoch": 0.136, "grad_norm": 2.046875, "learning_rate": 0.000212, "loss": 2.957, "mean_token_accuracy": 0.4387050449848175, "num_tokens": 187359.0, "step": 425 }, { "entropy": 3.104245185852051, "epoch": 0.1376, "grad_norm": 2.109375, "learning_rate": 0.0002145, "loss": 3.1075, "mean_token_accuracy": 0.411021426320076, "num_tokens": 189678.0, "step": 430 }, { "entropy": 3.119382953643799, "epoch": 0.1392, "grad_norm": 2.234375, "learning_rate": 0.00021700000000000002, "loss": 3.0925, "mean_token_accuracy": 0.41704494655132296, "num_tokens": 192025.0, "step": 435 }, { "entropy": 3.1301834106445314, "epoch": 0.1408, "grad_norm": 1.734375, "learning_rate": 0.0002195, "loss": 2.9788, "mean_token_accuracy": 0.4373115926980972, "num_tokens": 194266.0, "step": 440 }, { "entropy": 3.0479564666748047, "epoch": 0.1424, "grad_norm": 1.890625, "learning_rate": 0.000222, "loss": 2.9975, "mean_token_accuracy": 0.43155093789100646, "num_tokens": 196422.0, "step": 445 }, { "entropy": 3.0728238821029663, "epoch": 0.144, "grad_norm": 2.34375, "learning_rate": 0.0002245, "loss": 3.0344, "mean_token_accuracy": 0.42370045483112334, "num_tokens": 198724.0, "step": 450 }, { "entropy": 3.0693315267562866, "epoch": 0.1456, "grad_norm": 2.390625, "learning_rate": 0.00022700000000000002, "loss": 2.972, "mean_token_accuracy": 0.43325790464878083, "num_tokens": 200807.0, "step": 455 }, { "entropy": 3.021673321723938, "epoch": 0.1472, "grad_norm": 1.5390625, "learning_rate": 0.00022950000000000002, "loss": 2.9943, "mean_token_accuracy": 0.43570533096790315, "num_tokens": 203001.0, "step": 460 }, { "entropy": 3.0222464084625242, "epoch": 0.1488, "grad_norm": 1.4921875, "learning_rate": 0.00023200000000000003, "loss": 2.9737, "mean_token_accuracy": 0.4375020503997803, "num_tokens": 205375.0, "step": 465 }, { "entropy": 3.005727672576904, "epoch": 0.1504, "grad_norm": 2.515625, "learning_rate": 0.00023449999999999998, "loss": 3.0432, "mean_token_accuracy": 0.42280458509922025, "num_tokens": 207607.0, "step": 470 }, { "entropy": 3.023442602157593, "epoch": 0.152, "grad_norm": 1.609375, "learning_rate": 0.000237, "loss": 2.9931, "mean_token_accuracy": 0.43076890110969546, "num_tokens": 209759.0, "step": 475 }, { "entropy": 3.089386057853699, "epoch": 0.1536, "grad_norm": 2.34375, "learning_rate": 0.0002395, "loss": 2.971, "mean_token_accuracy": 0.44045292735099795, "num_tokens": 212021.0, "step": 480 }, { "entropy": 3.170826721191406, "epoch": 0.1552, "grad_norm": 1.59375, "learning_rate": 0.000242, "loss": 3.0953, "mean_token_accuracy": 0.41915816962718966, "num_tokens": 214649.0, "step": 485 }, { "entropy": 3.0228457689285277, "epoch": 0.1568, "grad_norm": 1.28125, "learning_rate": 0.0002445, "loss": 2.9822, "mean_token_accuracy": 0.4414661377668381, "num_tokens": 216914.0, "step": 490 }, { "entropy": 3.064147472381592, "epoch": 0.1584, "grad_norm": 1.5078125, "learning_rate": 0.000247, "loss": 2.9678, "mean_token_accuracy": 0.44378527998924255, "num_tokens": 219130.0, "step": 495 }, { "entropy": 2.927608871459961, "epoch": 0.16, "grad_norm": 1.78125, "learning_rate": 0.0002495, "loss": 2.9562, "mean_token_accuracy": 0.4438574850559235, "num_tokens": 221291.0, "step": 500 }, { "entropy": 2.995852160453796, "epoch": 0.1616, "grad_norm": 1.4375, "learning_rate": 0.000252, "loss": 2.9311, "mean_token_accuracy": 0.45227048397064207, "num_tokens": 223552.0, "step": 505 }, { "entropy": 3.032898449897766, "epoch": 0.1632, "grad_norm": 3.09375, "learning_rate": 0.0002545, "loss": 2.9778, "mean_token_accuracy": 0.4374843567609787, "num_tokens": 225808.0, "step": 510 }, { "entropy": 3.123756837844849, "epoch": 0.1648, "grad_norm": 1.671875, "learning_rate": 0.000257, "loss": 3.0563, "mean_token_accuracy": 0.4319914549589157, "num_tokens": 228247.0, "step": 515 }, { "entropy": 2.9924824237823486, "epoch": 0.1664, "grad_norm": 1.734375, "learning_rate": 0.0002595, "loss": 2.9754, "mean_token_accuracy": 0.44144312739372255, "num_tokens": 230415.0, "step": 520 }, { "entropy": 2.991943597793579, "epoch": 0.168, "grad_norm": 2.28125, "learning_rate": 0.000262, "loss": 2.9176, "mean_token_accuracy": 0.43671050667762756, "num_tokens": 232450.0, "step": 525 }, { "entropy": 3.020931434631348, "epoch": 0.1696, "grad_norm": 1.2421875, "learning_rate": 0.00026450000000000003, "loss": 2.9198, "mean_token_accuracy": 0.447659894824028, "num_tokens": 234523.0, "step": 530 }, { "entropy": 2.8841498374938963, "epoch": 0.1712, "grad_norm": 1.90625, "learning_rate": 0.00026700000000000004, "loss": 2.9091, "mean_token_accuracy": 0.4434284508228302, "num_tokens": 236512.0, "step": 535 }, { "entropy": 2.9701812505722045, "epoch": 0.1728, "grad_norm": 1.6640625, "learning_rate": 0.00026950000000000005, "loss": 2.9364, "mean_token_accuracy": 0.44666347801685335, "num_tokens": 238707.0, "step": 540 }, { "entropy": 3.0050160884857178, "epoch": 0.1744, "grad_norm": 2.03125, "learning_rate": 0.00027200000000000005, "loss": 3.0558, "mean_token_accuracy": 0.42309449315071107, "num_tokens": 241042.0, "step": 545 }, { "entropy": 3.164252257347107, "epoch": 0.176, "grad_norm": 2.6875, "learning_rate": 0.0002745, "loss": 3.014, "mean_token_accuracy": 0.43981266021728516, "num_tokens": 243391.0, "step": 550 }, { "entropy": 3.0667943716049195, "epoch": 0.1776, "grad_norm": 1.3671875, "learning_rate": 0.000277, "loss": 2.9544, "mean_token_accuracy": 0.43666306138038635, "num_tokens": 245479.0, "step": 555 }, { "entropy": 2.922476625442505, "epoch": 0.1792, "grad_norm": 1.71875, "learning_rate": 0.0002795, "loss": 2.9402, "mean_token_accuracy": 0.452865731716156, "num_tokens": 247586.0, "step": 560 }, { "entropy": 3.1145840883255005, "epoch": 0.1808, "grad_norm": 1.609375, "learning_rate": 0.00028199999999999997, "loss": 3.0212, "mean_token_accuracy": 0.4292920082807541, "num_tokens": 249962.0, "step": 565 }, { "entropy": 2.9589731216430666, "epoch": 0.1824, "grad_norm": 2.296875, "learning_rate": 0.0002845, "loss": 2.9843, "mean_token_accuracy": 0.4310429602861404, "num_tokens": 252104.0, "step": 570 }, { "entropy": 2.9423076152801513, "epoch": 0.184, "grad_norm": 1.5, "learning_rate": 0.000287, "loss": 2.9988, "mean_token_accuracy": 0.4464650273323059, "num_tokens": 254427.0, "step": 575 }, { "entropy": 3.0221834421157836, "epoch": 0.1856, "grad_norm": 1.3046875, "learning_rate": 0.0002895, "loss": 2.8728, "mean_token_accuracy": 0.45080505311489105, "num_tokens": 256509.0, "step": 580 }, { "entropy": 3.0010231256484987, "epoch": 0.1872, "grad_norm": 1.4921875, "learning_rate": 0.000292, "loss": 2.9455, "mean_token_accuracy": 0.4460518181324005, "num_tokens": 258594.0, "step": 585 }, { "entropy": 2.998606252670288, "epoch": 0.1888, "grad_norm": 1.390625, "learning_rate": 0.0002945, "loss": 3.031, "mean_token_accuracy": 0.42360666692256926, "num_tokens": 260979.0, "step": 590 }, { "entropy": 3.0627802133560182, "epoch": 0.1904, "grad_norm": 1.9140625, "learning_rate": 0.000297, "loss": 2.99, "mean_token_accuracy": 0.43044927418231965, "num_tokens": 263167.0, "step": 595 }, { "entropy": 3.1162281036376953, "epoch": 0.192, "grad_norm": 1.4296875, "learning_rate": 0.0002995, "loss": 3.0212, "mean_token_accuracy": 0.4319927155971527, "num_tokens": 265645.0, "step": 600 }, { "entropy": 2.969763422012329, "epoch": 0.1936, "grad_norm": 1.7265625, "learning_rate": 0.000302, "loss": 2.9369, "mean_token_accuracy": 0.44145881831645967, "num_tokens": 268014.0, "step": 605 }, { "entropy": 2.949857473373413, "epoch": 0.1952, "grad_norm": 1.90625, "learning_rate": 0.0003045, "loss": 3.0112, "mean_token_accuracy": 0.43535229861736296, "num_tokens": 270293.0, "step": 610 }, { "entropy": 2.989105558395386, "epoch": 0.1968, "grad_norm": 1.7578125, "learning_rate": 0.000307, "loss": 2.9482, "mean_token_accuracy": 0.4413189560174942, "num_tokens": 272525.0, "step": 615 }, { "entropy": 2.9540494680404663, "epoch": 0.1984, "grad_norm": 1.7265625, "learning_rate": 0.0003095, "loss": 2.9515, "mean_token_accuracy": 0.43812600076198577, "num_tokens": 274572.0, "step": 620 }, { "entropy": 2.9646738529205323, "epoch": 0.2, "grad_norm": 1.2421875, "learning_rate": 0.000312, "loss": 2.9532, "mean_token_accuracy": 0.4428831487894058, "num_tokens": 276698.0, "step": 625 }, { "entropy": 2.955666971206665, "epoch": 0.2016, "grad_norm": 1.4453125, "learning_rate": 0.0003145, "loss": 2.8801, "mean_token_accuracy": 0.4431019097566605, "num_tokens": 278700.0, "step": 630 }, { "entropy": 3.151052737236023, "epoch": 0.2032, "grad_norm": 1.6796875, "learning_rate": 0.000317, "loss": 2.9898, "mean_token_accuracy": 0.43132494390010834, "num_tokens": 280936.0, "step": 635 }, { "entropy": 3.023841953277588, "epoch": 0.2048, "grad_norm": 1.3828125, "learning_rate": 0.0003195, "loss": 3.0232, "mean_token_accuracy": 0.42879930734634397, "num_tokens": 283256.0, "step": 640 }, { "entropy": 3.039559078216553, "epoch": 0.2064, "grad_norm": 1.5078125, "learning_rate": 0.000322, "loss": 2.9623, "mean_token_accuracy": 0.4345048040151596, "num_tokens": 285570.0, "step": 645 }, { "entropy": 2.9134409427642822, "epoch": 0.208, "grad_norm": 1.5390625, "learning_rate": 0.00032450000000000003, "loss": 2.9873, "mean_token_accuracy": 0.43352285623550413, "num_tokens": 287823.0, "step": 650 }, { "entropy": 3.0298859596252443, "epoch": 0.2096, "grad_norm": 1.671875, "learning_rate": 0.00032700000000000003, "loss": 2.9864, "mean_token_accuracy": 0.4300517052412033, "num_tokens": 289954.0, "step": 655 }, { "entropy": 3.003275465965271, "epoch": 0.2112, "grad_norm": 1.3359375, "learning_rate": 0.00032950000000000004, "loss": 3.0211, "mean_token_accuracy": 0.4199120044708252, "num_tokens": 292251.0, "step": 660 }, { "entropy": 3.109331488609314, "epoch": 0.2128, "grad_norm": 1.359375, "learning_rate": 0.00033200000000000005, "loss": 2.9473, "mean_token_accuracy": 0.4423491328954697, "num_tokens": 294488.0, "step": 665 }, { "entropy": 2.987983465194702, "epoch": 0.2144, "grad_norm": 1.15625, "learning_rate": 0.00033450000000000005, "loss": 2.9821, "mean_token_accuracy": 0.4268669277429581, "num_tokens": 296761.0, "step": 670 }, { "entropy": 3.047322964668274, "epoch": 0.216, "grad_norm": 1.9609375, "learning_rate": 0.000337, "loss": 3.0138, "mean_token_accuracy": 0.42811892330646517, "num_tokens": 299053.0, "step": 675 }, { "entropy": 3.0243956565856935, "epoch": 0.2176, "grad_norm": 1.296875, "learning_rate": 0.0003395, "loss": 2.9843, "mean_token_accuracy": 0.4386782109737396, "num_tokens": 301318.0, "step": 680 }, { "entropy": 2.9484851121902467, "epoch": 0.2192, "grad_norm": 1.640625, "learning_rate": 0.000342, "loss": 2.9526, "mean_token_accuracy": 0.4427299976348877, "num_tokens": 303570.0, "step": 685 }, { "entropy": 2.970630574226379, "epoch": 0.2208, "grad_norm": 1.3671875, "learning_rate": 0.00034449999999999997, "loss": 2.9139, "mean_token_accuracy": 0.4486065596342087, "num_tokens": 305581.0, "step": 690 }, { "entropy": 2.907445216178894, "epoch": 0.2224, "grad_norm": 1.703125, "learning_rate": 0.000347, "loss": 2.9538, "mean_token_accuracy": 0.44807131588459015, "num_tokens": 307756.0, "step": 695 }, { "entropy": 2.9747052431106566, "epoch": 0.224, "grad_norm": 1.4140625, "learning_rate": 0.0003495, "loss": 2.9781, "mean_token_accuracy": 0.42667736411094664, "num_tokens": 309876.0, "step": 700 }, { "entropy": 3.0569634675979613, "epoch": 0.2256, "grad_norm": 1.3515625, "learning_rate": 0.000352, "loss": 2.9442, "mean_token_accuracy": 0.4365510940551758, "num_tokens": 312038.0, "step": 705 }, { "entropy": 2.9577913522720336, "epoch": 0.2272, "grad_norm": 1.1015625, "learning_rate": 0.0003545, "loss": 2.9492, "mean_token_accuracy": 0.4379378169775009, "num_tokens": 314222.0, "step": 710 }, { "entropy": 3.033062481880188, "epoch": 0.2288, "grad_norm": 1.4140625, "learning_rate": 0.000357, "loss": 3.0083, "mean_token_accuracy": 0.42992494702339173, "num_tokens": 316513.0, "step": 715 }, { "entropy": 2.8738722801208496, "epoch": 0.2304, "grad_norm": 1.671875, "learning_rate": 0.0003595, "loss": 2.8682, "mean_token_accuracy": 0.45183809399604796, "num_tokens": 318583.0, "step": 720 }, { "entropy": 2.9968027114868163, "epoch": 0.232, "grad_norm": 1.453125, "learning_rate": 0.000362, "loss": 3.0367, "mean_token_accuracy": 0.42545914351940156, "num_tokens": 320947.0, "step": 725 }, { "entropy": 2.881519889831543, "epoch": 0.2336, "grad_norm": 1.390625, "learning_rate": 0.0003645, "loss": 2.9511, "mean_token_accuracy": 0.4353248506784439, "num_tokens": 323015.0, "step": 730 }, { "entropy": 3.01561233997345, "epoch": 0.2352, "grad_norm": 1.2421875, "learning_rate": 0.000367, "loss": 2.9903, "mean_token_accuracy": 0.42550739645957947, "num_tokens": 325233.0, "step": 735 }, { "entropy": 2.789596676826477, "epoch": 0.2368, "grad_norm": 1.1796875, "learning_rate": 0.0003695, "loss": 2.8102, "mean_token_accuracy": 0.4636333823204041, "num_tokens": 327049.0, "step": 740 }, { "entropy": 3.01884069442749, "epoch": 0.2384, "grad_norm": 1.8984375, "learning_rate": 0.000372, "loss": 2.9087, "mean_token_accuracy": 0.44535877704620364, "num_tokens": 329121.0, "step": 745 }, { "entropy": 2.9490119218826294, "epoch": 0.24, "grad_norm": 1.640625, "learning_rate": 0.0003745, "loss": 2.94, "mean_token_accuracy": 0.4448482245206833, "num_tokens": 331291.0, "step": 750 }, { "entropy": 3.0040679693222048, "epoch": 0.2416, "grad_norm": 1.5234375, "learning_rate": 0.000377, "loss": 2.9627, "mean_token_accuracy": 0.447766450047493, "num_tokens": 333449.0, "step": 755 }, { "entropy": 2.9407071590423586, "epoch": 0.2432, "grad_norm": 1.21875, "learning_rate": 0.0003795, "loss": 2.8885, "mean_token_accuracy": 0.4578815370798111, "num_tokens": 335570.0, "step": 760 }, { "entropy": 3.020687961578369, "epoch": 0.2448, "grad_norm": 1.875, "learning_rate": 0.000382, "loss": 3.016, "mean_token_accuracy": 0.42489117980003355, "num_tokens": 337969.0, "step": 765 }, { "entropy": 2.959513711929321, "epoch": 0.2464, "grad_norm": 2.40625, "learning_rate": 0.0003845, "loss": 2.9832, "mean_token_accuracy": 0.43808636665344236, "num_tokens": 340246.0, "step": 770 }, { "entropy": 3.1199923038482664, "epoch": 0.248, "grad_norm": 1.671875, "learning_rate": 0.00038700000000000003, "loss": 3.0283, "mean_token_accuracy": 0.4226021945476532, "num_tokens": 342645.0, "step": 775 }, { "entropy": 3.157452654838562, "epoch": 0.2496, "grad_norm": 1.3046875, "learning_rate": 0.00038950000000000003, "loss": 2.9656, "mean_token_accuracy": 0.42893911004066465, "num_tokens": 344894.0, "step": 780 }, { "entropy": 2.969461464881897, "epoch": 0.2512, "grad_norm": 1.6796875, "learning_rate": 0.00039200000000000004, "loss": 2.9404, "mean_token_accuracy": 0.4340624839067459, "num_tokens": 346995.0, "step": 785 }, { "entropy": 3.068730115890503, "epoch": 0.2528, "grad_norm": 1.40625, "learning_rate": 0.00039450000000000005, "loss": 2.9592, "mean_token_accuracy": 0.4310222685337067, "num_tokens": 349259.0, "step": 790 }, { "entropy": 2.9265069007873534, "epoch": 0.2544, "grad_norm": 1.1796875, "learning_rate": 0.00039700000000000005, "loss": 2.8996, "mean_token_accuracy": 0.4491631776094437, "num_tokens": 351397.0, "step": 795 }, { "entropy": 3.026449418067932, "epoch": 0.256, "grad_norm": 1.2734375, "learning_rate": 0.0003995, "loss": 2.9379, "mean_token_accuracy": 0.44096632301807404, "num_tokens": 353522.0, "step": 800 }, { "entropy": 3.1140568971633913, "epoch": 0.2576, "grad_norm": 1.3515625, "learning_rate": 0.000402, "loss": 2.9859, "mean_token_accuracy": 0.429830738902092, "num_tokens": 355685.0, "step": 805 }, { "entropy": 2.9959977865219116, "epoch": 0.2592, "grad_norm": 1.5078125, "learning_rate": 0.0004045, "loss": 2.9359, "mean_token_accuracy": 0.4434337466955185, "num_tokens": 357788.0, "step": 810 }, { "entropy": 3.0590805292129515, "epoch": 0.2608, "grad_norm": 1.9609375, "learning_rate": 0.00040699999999999997, "loss": 2.921, "mean_token_accuracy": 0.4416261464357376, "num_tokens": 360034.0, "step": 815 }, { "entropy": 3.0206746816635133, "epoch": 0.2624, "grad_norm": 1.96875, "learning_rate": 0.0004095, "loss": 2.9952, "mean_token_accuracy": 0.4358320087194443, "num_tokens": 362337.0, "step": 820 }, { "entropy": 2.981063222885132, "epoch": 0.264, "grad_norm": 1.4921875, "learning_rate": 0.000412, "loss": 2.9855, "mean_token_accuracy": 0.4300475209951401, "num_tokens": 364656.0, "step": 825 }, { "entropy": 3.0999550104141234, "epoch": 0.2656, "grad_norm": 1.734375, "learning_rate": 0.0004145, "loss": 2.9506, "mean_token_accuracy": 0.4407676041126251, "num_tokens": 366750.0, "step": 830 }, { "entropy": 3.0867088556289675, "epoch": 0.2672, "grad_norm": 1.4765625, "learning_rate": 0.000417, "loss": 2.9583, "mean_token_accuracy": 0.444501668214798, "num_tokens": 369001.0, "step": 835 }, { "entropy": 3.077132511138916, "epoch": 0.2688, "grad_norm": 1.359375, "learning_rate": 0.0004195, "loss": 2.9721, "mean_token_accuracy": 0.4352314621210098, "num_tokens": 371317.0, "step": 840 }, { "entropy": 3.104713273048401, "epoch": 0.2704, "grad_norm": 1.2265625, "learning_rate": 0.000422, "loss": 2.9915, "mean_token_accuracy": 0.4379868686199188, "num_tokens": 373706.0, "step": 845 }, { "entropy": 3.0065540313720702, "epoch": 0.272, "grad_norm": 1.25, "learning_rate": 0.0004245, "loss": 2.9479, "mean_token_accuracy": 0.440329447388649, "num_tokens": 375839.0, "step": 850 }, { "entropy": 3.090069222450256, "epoch": 0.2736, "grad_norm": 1.5390625, "learning_rate": 0.000427, "loss": 2.9862, "mean_token_accuracy": 0.41951964795589447, "num_tokens": 378108.0, "step": 855 }, { "entropy": 3.1495770692825316, "epoch": 0.2752, "grad_norm": 1.6328125, "learning_rate": 0.0004295, "loss": 2.9677, "mean_token_accuracy": 0.4318382889032364, "num_tokens": 380321.0, "step": 860 }, { "entropy": 3.0006103992462156, "epoch": 0.2768, "grad_norm": 1.140625, "learning_rate": 0.000432, "loss": 2.9673, "mean_token_accuracy": 0.4359084039926529, "num_tokens": 382463.0, "step": 865 }, { "entropy": 2.9153977155685427, "epoch": 0.2784, "grad_norm": 1.234375, "learning_rate": 0.0004345, "loss": 2.9478, "mean_token_accuracy": 0.4496375352144241, "num_tokens": 384602.0, "step": 870 }, { "entropy": 2.993690919876099, "epoch": 0.28, "grad_norm": 1.375, "learning_rate": 0.000437, "loss": 2.993, "mean_token_accuracy": 0.4276667356491089, "num_tokens": 386978.0, "step": 875 }, { "entropy": 2.9283517837524413, "epoch": 0.2816, "grad_norm": 1.4453125, "learning_rate": 0.0004395, "loss": 2.9714, "mean_token_accuracy": 0.43373002409934996, "num_tokens": 389145.0, "step": 880 }, { "entropy": 2.9112048149108887, "epoch": 0.2832, "grad_norm": 1.265625, "learning_rate": 0.000442, "loss": 2.9148, "mean_token_accuracy": 0.4422976553440094, "num_tokens": 391202.0, "step": 885 }, { "entropy": 3.030206799507141, "epoch": 0.2848, "grad_norm": 1.3515625, "learning_rate": 0.0004445, "loss": 2.9844, "mean_token_accuracy": 0.43422403633594514, "num_tokens": 393516.0, "step": 890 }, { "entropy": 2.9839080333709718, "epoch": 0.2864, "grad_norm": 1.25, "learning_rate": 0.000447, "loss": 2.9402, "mean_token_accuracy": 0.4449074178934097, "num_tokens": 395624.0, "step": 895 }, { "entropy": 2.998187255859375, "epoch": 0.288, "grad_norm": 1.8515625, "learning_rate": 0.00044950000000000003, "loss": 2.9643, "mean_token_accuracy": 0.4357569754123688, "num_tokens": 397974.0, "step": 900 }, { "entropy": 3.0000478744506838, "epoch": 0.2896, "grad_norm": 1.3828125, "learning_rate": 0.00045200000000000004, "loss": 3.0394, "mean_token_accuracy": 0.41911774277687075, "num_tokens": 400354.0, "step": 905 }, { "entropy": 3.0431037664413454, "epoch": 0.2912, "grad_norm": 1.328125, "learning_rate": 0.00045450000000000004, "loss": 2.9779, "mean_token_accuracy": 0.43245829641819, "num_tokens": 402704.0, "step": 910 }, { "entropy": 3.117088961601257, "epoch": 0.2928, "grad_norm": 1.2109375, "learning_rate": 0.00045700000000000005, "loss": 3.0323, "mean_token_accuracy": 0.4208260804414749, "num_tokens": 405124.0, "step": 915 }, { "entropy": 3.0442737579345702, "epoch": 0.2944, "grad_norm": 1.7578125, "learning_rate": 0.00045950000000000006, "loss": 2.9502, "mean_token_accuracy": 0.4432258367538452, "num_tokens": 407402.0, "step": 920 }, { "entropy": 2.9706140041351317, "epoch": 0.296, "grad_norm": 1.2890625, "learning_rate": 0.000462, "loss": 2.9652, "mean_token_accuracy": 0.4415886074304581, "num_tokens": 409518.0, "step": 925 }, { "entropy": 3.0164210319519045, "epoch": 0.2976, "grad_norm": 1.3671875, "learning_rate": 0.0004645, "loss": 2.9788, "mean_token_accuracy": 0.4430856227874756, "num_tokens": 411754.0, "step": 930 }, { "entropy": 3.1146625757217405, "epoch": 0.2992, "grad_norm": 1.484375, "learning_rate": 0.000467, "loss": 2.9834, "mean_token_accuracy": 0.43455012440681456, "num_tokens": 414044.0, "step": 935 }, { "entropy": 2.9445823431015015, "epoch": 0.3008, "grad_norm": 1.2890625, "learning_rate": 0.0004695, "loss": 2.9046, "mean_token_accuracy": 0.4539070099592209, "num_tokens": 416088.0, "step": 940 }, { "entropy": 3.0215240478515626, "epoch": 0.3024, "grad_norm": 1.8984375, "learning_rate": 0.000472, "loss": 2.9676, "mean_token_accuracy": 0.433037930727005, "num_tokens": 418324.0, "step": 945 }, { "entropy": 2.9805382966995237, "epoch": 0.304, "grad_norm": 2.140625, "learning_rate": 0.0004745, "loss": 3.0229, "mean_token_accuracy": 0.42990840077400205, "num_tokens": 420764.0, "step": 950 }, { "entropy": 3.064919114112854, "epoch": 0.3056, "grad_norm": 1.3359375, "learning_rate": 0.000477, "loss": 2.9837, "mean_token_accuracy": 0.4254843920469284, "num_tokens": 423055.0, "step": 955 }, { "entropy": 3.0779817581176756, "epoch": 0.3072, "grad_norm": 1.2421875, "learning_rate": 0.0004795, "loss": 2.9578, "mean_token_accuracy": 0.43600958287715913, "num_tokens": 425298.0, "step": 960 }, { "entropy": 3.082110786437988, "epoch": 0.3088, "grad_norm": 1.46875, "learning_rate": 0.000482, "loss": 3.0336, "mean_token_accuracy": 0.4156427413225174, "num_tokens": 427666.0, "step": 965 }, { "entropy": 3.0057018995285034, "epoch": 0.3104, "grad_norm": 1.453125, "learning_rate": 0.0004845, "loss": 2.9635, "mean_token_accuracy": 0.4326969861984253, "num_tokens": 429950.0, "step": 970 }, { "entropy": 3.0246227979660034, "epoch": 0.312, "grad_norm": 1.2265625, "learning_rate": 0.000487, "loss": 2.8906, "mean_token_accuracy": 0.4557031333446503, "num_tokens": 431999.0, "step": 975 }, { "entropy": 3.0199164867401125, "epoch": 0.3136, "grad_norm": 1.1875, "learning_rate": 0.0004895, "loss": 2.9727, "mean_token_accuracy": 0.4296708106994629, "num_tokens": 434173.0, "step": 980 }, { "entropy": 3.036568593978882, "epoch": 0.3152, "grad_norm": 3.125, "learning_rate": 0.000492, "loss": 2.9299, "mean_token_accuracy": 0.4340952128171921, "num_tokens": 436336.0, "step": 985 }, { "entropy": 3.04384446144104, "epoch": 0.3168, "grad_norm": 1.1328125, "learning_rate": 0.0004945, "loss": 2.9661, "mean_token_accuracy": 0.4325526863336563, "num_tokens": 438604.0, "step": 990 }, { "entropy": 3.0534685850143433, "epoch": 0.3184, "grad_norm": 1.1953125, "learning_rate": 0.000497, "loss": 2.8914, "mean_token_accuracy": 0.4494841665029526, "num_tokens": 440698.0, "step": 995 }, { "entropy": 3.012888860702515, "epoch": 0.32, "grad_norm": 1.5078125, "learning_rate": 0.0004995, "loss": 3.004, "mean_token_accuracy": 0.418217259645462, "num_tokens": 442962.0, "step": 1000 }, { "entropy": 3.0333550453186033, "epoch": 0.3216, "grad_norm": 1.484375, "learning_rate": 0.000499998026082006, "loss": 3.0049, "mean_token_accuracy": 0.4294524550437927, "num_tokens": 445295.0, "step": 1005 }, { "entropy": 2.9813949346542357, "epoch": 0.3232, "grad_norm": 2.453125, "learning_rate": 0.0004999900070995136, "loss": 2.9546, "mean_token_accuracy": 0.4321258693933487, "num_tokens": 447449.0, "step": 1010 }, { "entropy": 2.964689517021179, "epoch": 0.3248, "grad_norm": 1.84375, "learning_rate": 0.0004999758199023239, "loss": 2.9786, "mean_token_accuracy": 0.437596806883812, "num_tokens": 449721.0, "step": 1015 }, { "entropy": 2.9920711755752563, "epoch": 0.3264, "grad_norm": 1.421875, "learning_rate": 0.0004999554648793858, "loss": 3.0115, "mean_token_accuracy": 0.4315613478422165, "num_tokens": 452199.0, "step": 1020 }, { "entropy": 3.005441904067993, "epoch": 0.328, "grad_norm": 1.2578125, "learning_rate": 0.0004999289425887425, "loss": 2.9995, "mean_token_accuracy": 0.42629573345184324, "num_tokens": 454363.0, "step": 1025 }, { "entropy": 3.0805040836334228, "epoch": 0.3296, "grad_norm": 1.28125, "learning_rate": 0.0004998962537575161, "loss": 2.9658, "mean_token_accuracy": 0.4282439470291138, "num_tokens": 456558.0, "step": 1030 }, { "entropy": 2.9859232187271116, "epoch": 0.3312, "grad_norm": 1.3359375, "learning_rate": 0.0004998573992818874, "loss": 2.9903, "mean_token_accuracy": 0.42836355566978457, "num_tokens": 458841.0, "step": 1035 }, { "entropy": 3.0495954751968384, "epoch": 0.3328, "grad_norm": 1.8046875, "learning_rate": 0.0004998123802270715, "loss": 2.9857, "mean_token_accuracy": 0.43424113988876345, "num_tokens": 461212.0, "step": 1040 }, { "entropy": 3.06514835357666, "epoch": 0.3344, "grad_norm": 1.1953125, "learning_rate": 0.0004997611978272886, "loss": 3.0313, "mean_token_accuracy": 0.4253895223140717, "num_tokens": 463538.0, "step": 1045 }, { "entropy": 3.028795576095581, "epoch": 0.336, "grad_norm": 1.1796875, "learning_rate": 0.0004997038534857298, "loss": 2.9737, "mean_token_accuracy": 0.4378691136837006, "num_tokens": 465904.0, "step": 1050 }, { "entropy": 3.001753878593445, "epoch": 0.3376, "grad_norm": 1.4453125, "learning_rate": 0.0004996403487745194, "loss": 2.9476, "mean_token_accuracy": 0.43258758187294005, "num_tokens": 468062.0, "step": 1055 }, { "entropy": 2.9926480770111086, "epoch": 0.3392, "grad_norm": 1.1875, "learning_rate": 0.000499570685434671, "loss": 2.869, "mean_token_accuracy": 0.45320001542568206, "num_tokens": 470129.0, "step": 1060 }, { "entropy": 3.029024934768677, "epoch": 0.3408, "grad_norm": 1.171875, "learning_rate": 0.0004994948653760405, "loss": 2.9471, "mean_token_accuracy": 0.4469340890645981, "num_tokens": 472245.0, "step": 1065 }, { "entropy": 3.0753472566604616, "epoch": 0.3424, "grad_norm": 1.171875, "learning_rate": 0.0004994128906772729, "loss": 2.955, "mean_token_accuracy": 0.4326744884252548, "num_tokens": 474467.0, "step": 1070 }, { "entropy": 3.0369976997375487, "epoch": 0.344, "grad_norm": 1.375, "learning_rate": 0.000499324763585746, "loss": 2.9612, "mean_token_accuracy": 0.43143860101699827, "num_tokens": 476660.0, "step": 1075 }, { "entropy": 3.019894027709961, "epoch": 0.3456, "grad_norm": 1.2890625, "learning_rate": 0.0004992304865175085, "loss": 2.9634, "mean_token_accuracy": 0.4395632088184357, "num_tokens": 478849.0, "step": 1080 }, { "entropy": 2.9239365100860595, "epoch": 0.3472, "grad_norm": 1.3046875, "learning_rate": 0.0004991300620572138, "loss": 2.9178, "mean_token_accuracy": 0.4388411432504654, "num_tokens": 480904.0, "step": 1085 }, { "entropy": 3.061780881881714, "epoch": 0.3488, "grad_norm": 1.53125, "learning_rate": 0.0004990234929580494, "loss": 3.0531, "mean_token_accuracy": 0.4172502398490906, "num_tokens": 483361.0, "step": 1090 }, { "entropy": 3.05807945728302, "epoch": 0.3504, "grad_norm": 1.3046875, "learning_rate": 0.0004989107821416609, "loss": 2.9711, "mean_token_accuracy": 0.42867847681045534, "num_tokens": 485576.0, "step": 1095 }, { "entropy": 2.8608323335647583, "epoch": 0.352, "grad_norm": 1.203125, "learning_rate": 0.0004987919326980723, "loss": 2.8963, "mean_token_accuracy": 0.4490462213754654, "num_tokens": 487656.0, "step": 1100 }, { "entropy": 2.9104605674743653, "epoch": 0.3536, "grad_norm": 1.15625, "learning_rate": 0.0004986669478856011, "loss": 2.9141, "mean_token_accuracy": 0.4411225587129593, "num_tokens": 489775.0, "step": 1105 }, { "entropy": 2.9821847677230835, "epoch": 0.3552, "grad_norm": 1.5, "learning_rate": 0.0004985358311307688, "loss": 2.9637, "mean_token_accuracy": 0.4325263828039169, "num_tokens": 492035.0, "step": 1110 }, { "entropy": 3.0659170866012575, "epoch": 0.3568, "grad_norm": 1.671875, "learning_rate": 0.0004983985860282081, "loss": 2.9379, "mean_token_accuracy": 0.43420119285583497, "num_tokens": 494307.0, "step": 1115 }, { "entropy": 2.891483497619629, "epoch": 0.3584, "grad_norm": 1.8203125, "learning_rate": 0.0004982552163405623, "loss": 2.9192, "mean_token_accuracy": 0.4528656393289566, "num_tokens": 496398.0, "step": 1120 }, { "entropy": 3.0359564304351805, "epoch": 0.36, "grad_norm": 1.140625, "learning_rate": 0.0004981057259983839, "loss": 3.1017, "mean_token_accuracy": 0.4107358932495117, "num_tokens": 498916.0, "step": 1125 }, { "entropy": 3.0484063148498537, "epoch": 0.3616, "grad_norm": 1.671875, "learning_rate": 0.0004979501191000262, "loss": 2.8949, "mean_token_accuracy": 0.4516554534435272, "num_tokens": 501085.0, "step": 1130 }, { "entropy": 2.8914608716964723, "epoch": 0.3632, "grad_norm": 1.2578125, "learning_rate": 0.0004977883999115311, "loss": 2.8729, "mean_token_accuracy": 0.4460421860218048, "num_tokens": 503182.0, "step": 1135 }, { "entropy": 2.944809412956238, "epoch": 0.3648, "grad_norm": 2.078125, "learning_rate": 0.0004976205728665113, "loss": 3.0037, "mean_token_accuracy": 0.43654546439647673, "num_tokens": 505524.0, "step": 1140 }, { "entropy": 3.1414292573928835, "epoch": 0.3664, "grad_norm": 1.46875, "learning_rate": 0.0004974466425660307, "loss": 3.0509, "mean_token_accuracy": 0.42192088067531586, "num_tokens": 507920.0, "step": 1145 }, { "entropy": 3.089212703704834, "epoch": 0.368, "grad_norm": 1.484375, "learning_rate": 0.0004972666137784759, "loss": 2.9637, "mean_token_accuracy": 0.42326963543891905, "num_tokens": 510175.0, "step": 1150 }, { "entropy": 3.0713165283203123, "epoch": 0.3696, "grad_norm": 1.2734375, "learning_rate": 0.0004970804914394271, "loss": 2.9515, "mean_token_accuracy": 0.43038663268089294, "num_tokens": 512352.0, "step": 1155 }, { "entropy": 3.1583834409713747, "epoch": 0.3712, "grad_norm": 1.1953125, "learning_rate": 0.0004968882806515225, "loss": 2.9301, "mean_token_accuracy": 0.4473735302686691, "num_tokens": 514540.0, "step": 1160 }, { "entropy": 3.032792258262634, "epoch": 0.3728, "grad_norm": 1.4453125, "learning_rate": 0.0004966899866843177, "loss": 2.899, "mean_token_accuracy": 0.44154741764068606, "num_tokens": 516724.0, "step": 1165 }, { "entropy": 2.9799720525741575, "epoch": 0.3744, "grad_norm": 1.328125, "learning_rate": 0.000496485614974142, "loss": 2.9398, "mean_token_accuracy": 0.4450356066226959, "num_tokens": 518959.0, "step": 1170 }, { "entropy": 3.011122298240662, "epoch": 0.376, "grad_norm": 1.625, "learning_rate": 0.0004962751711239492, "loss": 2.9025, "mean_token_accuracy": 0.4502583771944046, "num_tokens": 521145.0, "step": 1175 }, { "entropy": 2.9347607612609865, "epoch": 0.3776, "grad_norm": 1.2109375, "learning_rate": 0.0004960586609031636, "loss": 2.8704, "mean_token_accuracy": 0.4471834272146225, "num_tokens": 523202.0, "step": 1180 }, { "entropy": 3.0045668363571165, "epoch": 0.3792, "grad_norm": 2.8125, "learning_rate": 0.0004958360902475224, "loss": 2.9846, "mean_token_accuracy": 0.42844280898571013, "num_tokens": 525394.0, "step": 1185 }, { "entropy": 3.047668480873108, "epoch": 0.3808, "grad_norm": 1.375, "learning_rate": 0.0004956074652589125, "loss": 2.9535, "mean_token_accuracy": 0.43018776178359985, "num_tokens": 527636.0, "step": 1190 }, { "entropy": 3.0489161729812624, "epoch": 0.3824, "grad_norm": 1.0703125, "learning_rate": 0.0004953727922052035, "loss": 2.8932, "mean_token_accuracy": 0.45770248472690583, "num_tokens": 529725.0, "step": 1195 }, { "entropy": 2.9249891519546507, "epoch": 0.384, "grad_norm": 1.328125, "learning_rate": 0.0004951320775200756, "loss": 2.9604, "mean_token_accuracy": 0.4427888333797455, "num_tokens": 531944.0, "step": 1200 }, { "entropy": 2.9731184244155884, "epoch": 0.3856, "grad_norm": 1.8515625, "learning_rate": 0.0004948853278028436, "loss": 2.8919, "mean_token_accuracy": 0.4424555629491806, "num_tokens": 534064.0, "step": 1205 }, { "entropy": 3.0749728441238404, "epoch": 0.3872, "grad_norm": 1.234375, "learning_rate": 0.0004946325498182755, "loss": 2.9828, "mean_token_accuracy": 0.426044163107872, "num_tokens": 536472.0, "step": 1210 }, { "entropy": 3.007723021507263, "epoch": 0.3888, "grad_norm": 1.2421875, "learning_rate": 0.0004943737504964076, "loss": 2.9713, "mean_token_accuracy": 0.4258865177631378, "num_tokens": 538680.0, "step": 1215 }, { "entropy": 3.056945633888245, "epoch": 0.3904, "grad_norm": 1.15625, "learning_rate": 0.000494108936932354, "loss": 2.9573, "mean_token_accuracy": 0.4277519226074219, "num_tokens": 541059.0, "step": 1220 }, { "entropy": 2.9997201204299926, "epoch": 0.392, "grad_norm": 2.015625, "learning_rate": 0.0004938381163861124, "loss": 2.8825, "mean_token_accuracy": 0.4386983424425125, "num_tokens": 543188.0, "step": 1225 }, { "entropy": 2.9497172832489014, "epoch": 0.3936, "grad_norm": 1.4375, "learning_rate": 0.0004935612962823645, "loss": 2.9656, "mean_token_accuracy": 0.42417883276939394, "num_tokens": 545438.0, "step": 1230 }, { "entropy": 2.9807059526443482, "epoch": 0.3952, "grad_norm": 1.1484375, "learning_rate": 0.0004932784842102739, "loss": 2.8469, "mean_token_accuracy": 0.4599876910448074, "num_tokens": 547530.0, "step": 1235 }, { "entropy": 3.044364666938782, "epoch": 0.3968, "grad_norm": 1.109375, "learning_rate": 0.0004929896879232758, "loss": 2.9574, "mean_token_accuracy": 0.42144037783145905, "num_tokens": 549693.0, "step": 1240 }, { "entropy": 2.987114930152893, "epoch": 0.3984, "grad_norm": 1.2421875, "learning_rate": 0.0004926949153388668, "loss": 2.9397, "mean_token_accuracy": 0.44286964535713197, "num_tokens": 551742.0, "step": 1245 }, { "entropy": 3.1246228933334352, "epoch": 0.4, "grad_norm": 1.1015625, "learning_rate": 0.0004923941745383859, "loss": 2.9532, "mean_token_accuracy": 0.42484399378299714, "num_tokens": 553854.0, "step": 1250 }, { "entropy": 3.085980463027954, "epoch": 0.4016, "grad_norm": 1.15625, "learning_rate": 0.000492087473766794, "loss": 2.9982, "mean_token_accuracy": 0.42762279212474824, "num_tokens": 556241.0, "step": 1255 }, { "entropy": 2.9438645124435423, "epoch": 0.4032, "grad_norm": 1.4140625, "learning_rate": 0.000491774821432448, "loss": 2.9489, "mean_token_accuracy": 0.42370261549949645, "num_tokens": 558364.0, "step": 1260 }, { "entropy": 2.956605315208435, "epoch": 0.4048, "grad_norm": 1.3046875, "learning_rate": 0.0004914562261068693, "loss": 2.8865, "mean_token_accuracy": 0.45021391212940215, "num_tokens": 560575.0, "step": 1265 }, { "entropy": 2.972418260574341, "epoch": 0.4064, "grad_norm": 1.15625, "learning_rate": 0.0004911316965245098, "loss": 3.0046, "mean_token_accuracy": 0.4331658273935318, "num_tokens": 562916.0, "step": 1270 }, { "entropy": 2.883833312988281, "epoch": 0.408, "grad_norm": 1.234375, "learning_rate": 0.000490801241582512, "loss": 2.9411, "mean_token_accuracy": 0.4414670020341873, "num_tokens": 565033.0, "step": 1275 }, { "entropy": 3.0593276500701903, "epoch": 0.4096, "grad_norm": 1.1953125, "learning_rate": 0.000490464870340465, "loss": 2.9578, "mean_token_accuracy": 0.4305979788303375, "num_tokens": 567334.0, "step": 1280 }, { "entropy": 3.072508430480957, "epoch": 0.4112, "grad_norm": 1.171875, "learning_rate": 0.0004901225920201563, "loss": 2.9866, "mean_token_accuracy": 0.4319253474473953, "num_tokens": 569454.0, "step": 1285 }, { "entropy": 3.0835440397262572, "epoch": 0.4128, "grad_norm": 1.1796875, "learning_rate": 0.000489774416005319, "loss": 2.9481, "mean_token_accuracy": 0.43756221532821654, "num_tokens": 571745.0, "step": 1290 }, { "entropy": 3.0488495349884035, "epoch": 0.4144, "grad_norm": 1.5390625, "learning_rate": 0.0004894203518413742, "loss": 2.9714, "mean_token_accuracy": 0.43710216879844666, "num_tokens": 574038.0, "step": 1295 }, { "entropy": 3.094041681289673, "epoch": 0.416, "grad_norm": 1.4765625, "learning_rate": 0.0004890604092351701, "loss": 2.9532, "mean_token_accuracy": 0.4336225062608719, "num_tokens": 576285.0, "step": 1300 }, { "entropy": 3.0341554403305055, "epoch": 0.4176, "grad_norm": 1.09375, "learning_rate": 0.000488694598054715, "loss": 2.923, "mean_token_accuracy": 0.4342182636260986, "num_tokens": 578273.0, "step": 1305 }, { "entropy": 3.1518479347229005, "epoch": 0.4192, "grad_norm": 1.421875, "learning_rate": 0.0004883229283289071, "loss": 2.919, "mean_token_accuracy": 0.4374409466981888, "num_tokens": 580398.0, "step": 1310 }, { "entropy": 3.0593973636627196, "epoch": 0.4208, "grad_norm": 2.84375, "learning_rate": 0.00048794541024725993, "loss": 2.9321, "mean_token_accuracy": 0.43606520593166354, "num_tokens": 582612.0, "step": 1315 }, { "entropy": 2.961854672431946, "epoch": 0.4224, "grad_norm": 1.234375, "learning_rate": 0.0004875620541596221, "loss": 2.925, "mean_token_accuracy": 0.4340157449245453, "num_tokens": 584750.0, "step": 1320 }, { "entropy": 2.9912842750549316, "epoch": 0.424, "grad_norm": 1.1484375, "learning_rate": 0.00048717287057589454, "loss": 2.9522, "mean_token_accuracy": 0.43737233281135557, "num_tokens": 586829.0, "step": 1325 }, { "entropy": 3.086051678657532, "epoch": 0.4256, "grad_norm": 1.015625, "learning_rate": 0.0004867778701657417, "loss": 2.9892, "mean_token_accuracy": 0.4262147516012192, "num_tokens": 589289.0, "step": 1330 }, { "entropy": 3.0404081106185914, "epoch": 0.4272, "grad_norm": 1.40625, "learning_rate": 0.00048637706375829955, "loss": 2.8758, "mean_token_accuracy": 0.44268335700035094, "num_tokens": 591302.0, "step": 1335 }, { "entropy": 3.023822283744812, "epoch": 0.4288, "grad_norm": 1.265625, "learning_rate": 0.000485970462341878, "loss": 2.915, "mean_token_accuracy": 0.43806929886341095, "num_tokens": 593388.0, "step": 1340 }, { "entropy": 2.9713682651519777, "epoch": 0.4304, "grad_norm": 11.4375, "learning_rate": 0.00048555807706366044, "loss": 2.9538, "mean_token_accuracy": 0.4454095005989075, "num_tokens": 595583.0, "step": 1345 }, { "entropy": 2.984753680229187, "epoch": 0.432, "grad_norm": 1.65625, "learning_rate": 0.00048513991922939756, "loss": 2.9307, "mean_token_accuracy": 0.45300622284412384, "num_tokens": 597649.0, "step": 1350 }, { "entropy": 3.106572389602661, "epoch": 0.4336, "grad_norm": 1.171875, "learning_rate": 0.00048471600030309744, "loss": 3.0381, "mean_token_accuracy": 0.41848199963569643, "num_tokens": 600165.0, "step": 1355 }, { "entropy": 3.0415312528610228, "epoch": 0.4352, "grad_norm": 1.1484375, "learning_rate": 0.00048428633190671186, "loss": 2.926, "mean_token_accuracy": 0.447231525182724, "num_tokens": 602276.0, "step": 1360 }, { "entropy": 3.0783979177474974, "epoch": 0.4368, "grad_norm": 1.140625, "learning_rate": 0.0004838509258198167, "loss": 2.9372, "mean_token_accuracy": 0.44867193400859834, "num_tokens": 604593.0, "step": 1365 }, { "entropy": 2.9634838819503786, "epoch": 0.4384, "grad_norm": 1.4375, "learning_rate": 0.00048340979397929, "loss": 2.9005, "mean_token_accuracy": 0.4389187514781952, "num_tokens": 606676.0, "step": 1370 }, { "entropy": 2.9471078634262087, "epoch": 0.44, "grad_norm": 1.46875, "learning_rate": 0.00048296294847898386, "loss": 2.9548, "mean_token_accuracy": 0.43695137202739714, "num_tokens": 608911.0, "step": 1375 }, { "entropy": 2.9131209373474123, "epoch": 0.4416, "grad_norm": 1.0390625, "learning_rate": 0.0004825104015693934, "loss": 2.9079, "mean_token_accuracy": 0.44549469351768495, "num_tokens": 611110.0, "step": 1380 }, { "entropy": 2.9700817823410035, "epoch": 0.4432, "grad_norm": 1.0625, "learning_rate": 0.0004820521656573208, "loss": 2.9158, "mean_token_accuracy": 0.4435183942317963, "num_tokens": 613283.0, "step": 1385 }, { "entropy": 2.939583492279053, "epoch": 0.4448, "grad_norm": 1.1171875, "learning_rate": 0.00048158825330553505, "loss": 2.877, "mean_token_accuracy": 0.4486056208610535, "num_tokens": 615326.0, "step": 1390 }, { "entropy": 2.9285218000411986, "epoch": 0.4464, "grad_norm": 1.1796875, "learning_rate": 0.00048111867723242763, "loss": 2.9054, "mean_token_accuracy": 0.44363516569137573, "num_tokens": 617370.0, "step": 1395 }, { "entropy": 2.9988037586212157, "epoch": 0.448, "grad_norm": 1.0703125, "learning_rate": 0.0004806434503116637, "loss": 2.91, "mean_token_accuracy": 0.44267665445804594, "num_tokens": 619738.0, "step": 1400 }, { "entropy": 2.9457010507583616, "epoch": 0.4496, "grad_norm": 1.390625, "learning_rate": 0.0004801625855718296, "loss": 2.9404, "mean_token_accuracy": 0.442170450091362, "num_tokens": 622062.0, "step": 1405 }, { "entropy": 2.8639967918395994, "epoch": 0.4512, "grad_norm": 1.4296875, "learning_rate": 0.00047967609619607477, "loss": 2.8822, "mean_token_accuracy": 0.44169872999191284, "num_tokens": 624076.0, "step": 1410 }, { "entropy": 2.9699059247970583, "epoch": 0.4528, "grad_norm": 1.453125, "learning_rate": 0.0004791839955217513, "loss": 2.9397, "mean_token_accuracy": 0.4383286744356155, "num_tokens": 626312.0, "step": 1415 }, { "entropy": 3.0286834955215456, "epoch": 0.4544, "grad_norm": 1.5078125, "learning_rate": 0.00047868629704004786, "loss": 2.8866, "mean_token_accuracy": 0.4450020492076874, "num_tokens": 628420.0, "step": 1420 }, { "entropy": 2.930131721496582, "epoch": 0.456, "grad_norm": 1.2109375, "learning_rate": 0.00047818301439561965, "loss": 2.8403, "mean_token_accuracy": 0.45827619135379793, "num_tokens": 630369.0, "step": 1425 }, { "entropy": 2.9265647172927856, "epoch": 0.4576, "grad_norm": 1.6953125, "learning_rate": 0.00047767416138621454, "loss": 2.9362, "mean_token_accuracy": 0.4424218088388443, "num_tokens": 632554.0, "step": 1430 }, { "entropy": 3.0317028760910034, "epoch": 0.4592, "grad_norm": 1.0703125, "learning_rate": 0.000477159751962295, "loss": 2.9333, "mean_token_accuracy": 0.4429584532976151, "num_tokens": 634699.0, "step": 1435 }, { "entropy": 2.8802489757537844, "epoch": 0.4608, "grad_norm": 1.0703125, "learning_rate": 0.00047663980022665507, "loss": 2.8158, "mean_token_accuracy": 0.4581249088048935, "num_tokens": 636651.0, "step": 1440 }, { "entropy": 2.867780828475952, "epoch": 0.4624, "grad_norm": 1.1328125, "learning_rate": 0.00047611432043403437, "loss": 2.9079, "mean_token_accuracy": 0.4447682827711105, "num_tokens": 638840.0, "step": 1445 }, { "entropy": 3.0133840322494505, "epoch": 0.464, "grad_norm": 1.3125, "learning_rate": 0.0004755833269907267, "loss": 2.9293, "mean_token_accuracy": 0.4471138477325439, "num_tokens": 641071.0, "step": 1450 }, { "entropy": 3.0760305643081667, "epoch": 0.4656, "grad_norm": 1.3203125, "learning_rate": 0.0004750468344541857, "loss": 2.9671, "mean_token_accuracy": 0.43633088767528533, "num_tokens": 643260.0, "step": 1455 }, { "entropy": 3.037857985496521, "epoch": 0.4672, "grad_norm": 1.0703125, "learning_rate": 0.00047450485753262525, "loss": 2.9627, "mean_token_accuracy": 0.4231949210166931, "num_tokens": 645564.0, "step": 1460 }, { "entropy": 3.0435588121414185, "epoch": 0.4688, "grad_norm": 1.1640625, "learning_rate": 0.00047395741108461633, "loss": 2.9523, "mean_token_accuracy": 0.43553089797496797, "num_tokens": 647788.0, "step": 1465 }, { "entropy": 3.033423590660095, "epoch": 0.4704, "grad_norm": 1.0625, "learning_rate": 0.00047340451011867985, "loss": 2.9619, "mean_token_accuracy": 0.4316022783517838, "num_tokens": 650215.0, "step": 1470 }, { "entropy": 3.027427577972412, "epoch": 0.472, "grad_norm": 1.0703125, "learning_rate": 0.00047284616979287515, "loss": 2.9588, "mean_token_accuracy": 0.43038985729217527, "num_tokens": 652780.0, "step": 1475 }, { "entropy": 2.8937772512435913, "epoch": 0.4736, "grad_norm": 1.1796875, "learning_rate": 0.00047228240541438433, "loss": 2.9182, "mean_token_accuracy": 0.4420487999916077, "num_tokens": 654946.0, "step": 1480 }, { "entropy": 2.9493401527404783, "epoch": 0.4752, "grad_norm": 1.15625, "learning_rate": 0.00047171323243909257, "loss": 2.9034, "mean_token_accuracy": 0.44070430397987365, "num_tokens": 657090.0, "step": 1485 }, { "entropy": 3.005044174194336, "epoch": 0.4768, "grad_norm": 1.109375, "learning_rate": 0.00047113866647116457, "loss": 2.9418, "mean_token_accuracy": 0.43523562550544737, "num_tokens": 659172.0, "step": 1490 }, { "entropy": 3.0409275770187376, "epoch": 0.4784, "grad_norm": 1.3828125, "learning_rate": 0.0004705587232626164, "loss": 2.9928, "mean_token_accuracy": 0.43100384473800657, "num_tokens": 661504.0, "step": 1495 }, { "entropy": 3.029221296310425, "epoch": 0.48, "grad_norm": 1.1875, "learning_rate": 0.00046997341871288424, "loss": 2.9324, "mean_token_accuracy": 0.43912685811519625, "num_tokens": 663742.0, "step": 1500 }, { "entropy": 3.0940512657165526, "epoch": 0.4816, "grad_norm": 1.1328125, "learning_rate": 0.0004693827688683879, "loss": 2.9814, "mean_token_accuracy": 0.42807449996471403, "num_tokens": 666153.0, "step": 1505 }, { "entropy": 2.901436519622803, "epoch": 0.4832, "grad_norm": 1.6796875, "learning_rate": 0.0004687867899220914, "loss": 2.8255, "mean_token_accuracy": 0.45105704963207244, "num_tokens": 668061.0, "step": 1510 }, { "entropy": 2.985918402671814, "epoch": 0.4848, "grad_norm": 1.40625, "learning_rate": 0.00046818549821305846, "loss": 2.9642, "mean_token_accuracy": 0.43088276088237765, "num_tokens": 670419.0, "step": 1515 }, { "entropy": 2.9436598300933836, "epoch": 0.4864, "grad_norm": 1.1328125, "learning_rate": 0.00046757891022600494, "loss": 2.9166, "mean_token_accuracy": 0.4460465252399445, "num_tokens": 672539.0, "step": 1520 }, { "entropy": 2.915910744667053, "epoch": 0.488, "grad_norm": 1.265625, "learning_rate": 0.0004669670425908471, "loss": 2.8774, "mean_token_accuracy": 0.44755939245223997, "num_tokens": 674591.0, "step": 1525 }, { "entropy": 3.003065752983093, "epoch": 0.4896, "grad_norm": 1.1953125, "learning_rate": 0.0004663499120822451, "loss": 2.9644, "mean_token_accuracy": 0.440701425075531, "num_tokens": 676917.0, "step": 1530 }, { "entropy": 2.992565131187439, "epoch": 0.4912, "grad_norm": 1.5, "learning_rate": 0.0004657275356191437, "loss": 2.924, "mean_token_accuracy": 0.4346348136663437, "num_tokens": 679114.0, "step": 1535 }, { "entropy": 2.9336121797561647, "epoch": 0.4928, "grad_norm": 1.28125, "learning_rate": 0.00046509993026430804, "loss": 2.9084, "mean_token_accuracy": 0.43787922263145446, "num_tokens": 681359.0, "step": 1540 }, { "entropy": 2.8388763189315798, "epoch": 0.4944, "grad_norm": 1.421875, "learning_rate": 0.0004644671132238558, "loss": 2.8432, "mean_token_accuracy": 0.44864566922187804, "num_tokens": 683432.0, "step": 1545 }, { "entropy": 2.927882671356201, "epoch": 0.496, "grad_norm": 1.140625, "learning_rate": 0.00046382910184678585, "loss": 2.9327, "mean_token_accuracy": 0.4384798973798752, "num_tokens": 685759.0, "step": 1550 }, { "entropy": 2.9877994775772097, "epoch": 0.4976, "grad_norm": 1.1640625, "learning_rate": 0.0004631859136245025, "loss": 2.9052, "mean_token_accuracy": 0.4395827084779739, "num_tokens": 688034.0, "step": 1555 }, { "entropy": 2.9806543827056884, "epoch": 0.4992, "grad_norm": 1.203125, "learning_rate": 0.0004625375661903357, "loss": 2.9182, "mean_token_accuracy": 0.4401173233985901, "num_tokens": 690222.0, "step": 1560 }, { "entropy": 3.036587643623352, "epoch": 0.5008, "grad_norm": 1.3515625, "learning_rate": 0.00046188407731905787, "loss": 3.055, "mean_token_accuracy": 0.41269941329956056, "num_tokens": 692836.0, "step": 1565 }, { "entropy": 2.9350876331329347, "epoch": 0.5024, "grad_norm": 1.9453125, "learning_rate": 0.00046122546492639643, "loss": 2.8418, "mean_token_accuracy": 0.4560267567634583, "num_tokens": 694824.0, "step": 1570 }, { "entropy": 2.887144947052002, "epoch": 0.504, "grad_norm": 1.3359375, "learning_rate": 0.000460561747068543, "loss": 2.8881, "mean_token_accuracy": 0.44894916117191314, "num_tokens": 696887.0, "step": 1575 }, { "entropy": 2.946301221847534, "epoch": 0.5056, "grad_norm": 1.375, "learning_rate": 0.0004598929419416578, "loss": 2.9857, "mean_token_accuracy": 0.42084073424339297, "num_tokens": 699295.0, "step": 1580 }, { "entropy": 2.9399414539337156, "epoch": 0.5072, "grad_norm": 1.28125, "learning_rate": 0.00045921906788137123, "loss": 2.9725, "mean_token_accuracy": 0.4304078370332718, "num_tokens": 701461.0, "step": 1585 }, { "entropy": 2.9465761184692383, "epoch": 0.5088, "grad_norm": 1.265625, "learning_rate": 0.00045854014336228115, "loss": 2.931, "mean_token_accuracy": 0.4429843842983246, "num_tokens": 703700.0, "step": 1590 }, { "entropy": 2.950738787651062, "epoch": 0.5104, "grad_norm": 1.140625, "learning_rate": 0.00045785618699744615, "loss": 2.8607, "mean_token_accuracy": 0.45249311029911043, "num_tokens": 705769.0, "step": 1595 }, { "entropy": 2.9854349851608277, "epoch": 0.512, "grad_norm": 1.0703125, "learning_rate": 0.00045716721753787543, "loss": 2.9489, "mean_token_accuracy": 0.43289425075054166, "num_tokens": 708060.0, "step": 1600 }, { "entropy": 2.9239821672439574, "epoch": 0.5136, "grad_norm": 1.1484375, "learning_rate": 0.0004564732538720148, "loss": 2.8748, "mean_token_accuracy": 0.4475723296403885, "num_tokens": 710198.0, "step": 1605 }, { "entropy": 2.9703863859176636, "epoch": 0.5152, "grad_norm": 1.15625, "learning_rate": 0.00045577431502522877, "loss": 2.9549, "mean_token_accuracy": 0.44239388704299926, "num_tokens": 712495.0, "step": 1610 }, { "entropy": 2.939270782470703, "epoch": 0.5168, "grad_norm": 1.3125, "learning_rate": 0.0004550704201592787, "loss": 2.9157, "mean_token_accuracy": 0.4415583312511444, "num_tokens": 714717.0, "step": 1615 }, { "entropy": 2.943677377700806, "epoch": 0.5184, "grad_norm": 1.078125, "learning_rate": 0.0004543615885717981, "loss": 2.8977, "mean_token_accuracy": 0.44517011046409605, "num_tokens": 716825.0, "step": 1620 }, { "entropy": 2.9194502353668215, "epoch": 0.52, "grad_norm": 1.109375, "learning_rate": 0.00045364783969576296, "loss": 2.9083, "mean_token_accuracy": 0.44007180333137513, "num_tokens": 719108.0, "step": 1625 }, { "entropy": 2.8892699480056763, "epoch": 0.5216, "grad_norm": 8.9375, "learning_rate": 0.0004529291930989592, "loss": 2.9274, "mean_token_accuracy": 0.43131971955299375, "num_tokens": 721440.0, "step": 1630 }, { "entropy": 2.934997797012329, "epoch": 0.5232, "grad_norm": 1.5390625, "learning_rate": 0.0004522056684834464, "loss": 2.9128, "mean_token_accuracy": 0.43331984579563143, "num_tokens": 723564.0, "step": 1635 }, { "entropy": 2.9300397396087647, "epoch": 0.5248, "grad_norm": 1.125, "learning_rate": 0.0004514772856850173, "loss": 2.9244, "mean_token_accuracy": 0.42689967453479766, "num_tokens": 725666.0, "step": 1640 }, { "entropy": 3.0150431632995605, "epoch": 0.5264, "grad_norm": 1.2890625, "learning_rate": 0.0004507440646726542, "loss": 2.8946, "mean_token_accuracy": 0.4354646295309067, "num_tokens": 727834.0, "step": 1645 }, { "entropy": 2.992683982849121, "epoch": 0.528, "grad_norm": 1.140625, "learning_rate": 0.0004500060255479818, "loss": 2.9717, "mean_token_accuracy": 0.4252068758010864, "num_tokens": 730084.0, "step": 1650 }, { "entropy": 2.9637508153915406, "epoch": 0.5296, "grad_norm": 1.2265625, "learning_rate": 0.0004492631885447151, "loss": 2.8834, "mean_token_accuracy": 0.4451540172100067, "num_tokens": 732171.0, "step": 1655 }, { "entropy": 2.958834481239319, "epoch": 0.5312, "grad_norm": 1.2265625, "learning_rate": 0.00044851557402810616, "loss": 2.9466, "mean_token_accuracy": 0.43138331472873687, "num_tokens": 734415.0, "step": 1660 }, { "entropy": 2.9554142236709593, "epoch": 0.5328, "grad_norm": 0.98828125, "learning_rate": 0.00044776320249438444, "loss": 2.935, "mean_token_accuracy": 0.4441295176744461, "num_tokens": 736609.0, "step": 1665 }, { "entropy": 2.9917065143585204, "epoch": 0.5344, "grad_norm": 1.171875, "learning_rate": 0.00044700609457019565, "loss": 2.9069, "mean_token_accuracy": 0.4476753741502762, "num_tokens": 738826.0, "step": 1670 }, { "entropy": 2.9505895614624023, "epoch": 0.536, "grad_norm": 1.2578125, "learning_rate": 0.0004462442710120359, "loss": 2.9064, "mean_token_accuracy": 0.44069992899894717, "num_tokens": 740953.0, "step": 1675 }, { "entropy": 2.8723142385482787, "epoch": 0.5376, "grad_norm": 1.140625, "learning_rate": 0.000445477752705683, "loss": 2.8871, "mean_token_accuracy": 0.4433819383382797, "num_tokens": 743108.0, "step": 1680 }, { "entropy": 2.9634553670883177, "epoch": 0.5392, "grad_norm": 1.140625, "learning_rate": 0.00044470656066562336, "loss": 2.9345, "mean_token_accuracy": 0.440775528550148, "num_tokens": 745396.0, "step": 1685 }, { "entropy": 2.9189784049987795, "epoch": 0.5408, "grad_norm": 1.0859375, "learning_rate": 0.0004439307160344765, "loss": 2.9224, "mean_token_accuracy": 0.4445586293935776, "num_tokens": 747716.0, "step": 1690 }, { "entropy": 2.9116557359695436, "epoch": 0.5424, "grad_norm": 1.1640625, "learning_rate": 0.00044315024008241473, "loss": 2.889, "mean_token_accuracy": 0.4398995012044907, "num_tokens": 749793.0, "step": 1695 }, { "entropy": 2.949750542640686, "epoch": 0.544, "grad_norm": 1.3125, "learning_rate": 0.0004423651542065806, "loss": 2.9601, "mean_token_accuracy": 0.430759871006012, "num_tokens": 752081.0, "step": 1700 }, { "entropy": 2.927095341682434, "epoch": 0.5456, "grad_norm": 1.078125, "learning_rate": 0.00044157547993050006, "loss": 2.9689, "mean_token_accuracy": 0.43347471952438354, "num_tokens": 754343.0, "step": 1705 }, { "entropy": 3.0887319087982177, "epoch": 0.5472, "grad_norm": 1.0390625, "learning_rate": 0.00044078123890349227, "loss": 3.014, "mean_token_accuracy": 0.42509621381759644, "num_tokens": 756790.0, "step": 1710 }, { "entropy": 3.0619884014129637, "epoch": 0.5488, "grad_norm": 1.1953125, "learning_rate": 0.00043998245290007606, "loss": 2.9685, "mean_token_accuracy": 0.4327192634344101, "num_tokens": 759036.0, "step": 1715 }, { "entropy": 2.9572569608688353, "epoch": 0.5504, "grad_norm": 1.125, "learning_rate": 0.00043917914381937323, "loss": 2.8924, "mean_token_accuracy": 0.4381801664829254, "num_tokens": 761222.0, "step": 1720 }, { "entropy": 2.9272555351257323, "epoch": 0.552, "grad_norm": 1.109375, "learning_rate": 0.00043837133368450815, "loss": 2.9147, "mean_token_accuracy": 0.44456625580787656, "num_tokens": 763390.0, "step": 1725 }, { "entropy": 2.919555687904358, "epoch": 0.5536, "grad_norm": 1.1015625, "learning_rate": 0.0004375590446420037, "loss": 2.856, "mean_token_accuracy": 0.45251292884349825, "num_tokens": 765395.0, "step": 1730 }, { "entropy": 2.9558494806289675, "epoch": 0.5552, "grad_norm": 1.078125, "learning_rate": 0.0004367422989611743, "loss": 2.9136, "mean_token_accuracy": 0.44255523979663847, "num_tokens": 767589.0, "step": 1735 }, { "entropy": 2.9474855422973634, "epoch": 0.5568, "grad_norm": 1.3203125, "learning_rate": 0.0004359211190335153, "loss": 2.8724, "mean_token_accuracy": 0.44648954570293425, "num_tokens": 769670.0, "step": 1740 }, { "entropy": 2.904417037963867, "epoch": 0.5584, "grad_norm": 1.15625, "learning_rate": 0.00043509552737208923, "loss": 2.9044, "mean_token_accuracy": 0.44372026026248934, "num_tokens": 771896.0, "step": 1745 }, { "entropy": 2.8864651679992677, "epoch": 0.56, "grad_norm": 1.4765625, "learning_rate": 0.00043426554661090853, "loss": 2.8858, "mean_token_accuracy": 0.44257611930370333, "num_tokens": 774003.0, "step": 1750 }, { "entropy": 2.872239685058594, "epoch": 0.5616, "grad_norm": 1.1796875, "learning_rate": 0.00043343119950431516, "loss": 2.8676, "mean_token_accuracy": 0.4535475343465805, "num_tokens": 776137.0, "step": 1755 }, { "entropy": 2.884163475036621, "epoch": 0.5632, "grad_norm": 1.125, "learning_rate": 0.00043259250892635644, "loss": 2.9085, "mean_token_accuracy": 0.4429836988449097, "num_tokens": 778347.0, "step": 1760 }, { "entropy": 2.9461241483688356, "epoch": 0.5648, "grad_norm": 1.1796875, "learning_rate": 0.0004317494978701582, "loss": 2.934, "mean_token_accuracy": 0.44210954308509826, "num_tokens": 780476.0, "step": 1765 }, { "entropy": 3.0180752277374268, "epoch": 0.5664, "grad_norm": 2.21875, "learning_rate": 0.0004309021894472943, "loss": 2.9508, "mean_token_accuracy": 0.44288161396980286, "num_tokens": 782669.0, "step": 1770 }, { "entropy": 2.996902656555176, "epoch": 0.568, "grad_norm": 1.1640625, "learning_rate": 0.0004300506068871534, "loss": 2.8806, "mean_token_accuracy": 0.44376530051231383, "num_tokens": 784801.0, "step": 1775 }, { "entropy": 2.9082651138305664, "epoch": 0.5696, "grad_norm": 1.28125, "learning_rate": 0.00042919477353630135, "loss": 2.8948, "mean_token_accuracy": 0.44319796562194824, "num_tokens": 787037.0, "step": 1780 }, { "entropy": 2.8395618438720702, "epoch": 0.5712, "grad_norm": 1.2890625, "learning_rate": 0.000428334712857842, "loss": 2.9198, "mean_token_accuracy": 0.4346603661775589, "num_tokens": 789172.0, "step": 1785 }, { "entropy": 2.956600213050842, "epoch": 0.5728, "grad_norm": 1.5859375, "learning_rate": 0.00042747044843077304, "loss": 2.8649, "mean_token_accuracy": 0.447929722070694, "num_tokens": 791265.0, "step": 1790 }, { "entropy": 2.9591255664825438, "epoch": 0.5744, "grad_norm": 0.98828125, "learning_rate": 0.00042660200394934047, "loss": 2.8336, "mean_token_accuracy": 0.4566278040409088, "num_tokens": 793366.0, "step": 1795 }, { "entropy": 2.945344161987305, "epoch": 0.576, "grad_norm": 1.0625, "learning_rate": 0.00042572940322238844, "loss": 2.8711, "mean_token_accuracy": 0.4500783771276474, "num_tokens": 795450.0, "step": 1800 }, { "entropy": 2.9862860679626464, "epoch": 0.5776, "grad_norm": 1.140625, "learning_rate": 0.00042485267017270664, "loss": 2.9724, "mean_token_accuracy": 0.4373950719833374, "num_tokens": 797639.0, "step": 1805 }, { "entropy": 2.900651478767395, "epoch": 0.5792, "grad_norm": 1.109375, "learning_rate": 0.0004239718288363745, "loss": 2.9278, "mean_token_accuracy": 0.44675204157829285, "num_tokens": 799715.0, "step": 1810 }, { "entropy": 3.092992377281189, "epoch": 0.5808, "grad_norm": 1.25, "learning_rate": 0.0004230869033621023, "loss": 2.9254, "mean_token_accuracy": 0.438578274846077, "num_tokens": 801909.0, "step": 1815 }, { "entropy": 3.009065556526184, "epoch": 0.5824, "grad_norm": 1.1875, "learning_rate": 0.0004221979180105688, "loss": 2.8729, "mean_token_accuracy": 0.4487992852926254, "num_tokens": 803925.0, "step": 1820 }, { "entropy": 2.987674331665039, "epoch": 0.584, "grad_norm": 1.1484375, "learning_rate": 0.00042130489715375645, "loss": 2.8953, "mean_token_accuracy": 0.44306851327419283, "num_tokens": 806131.0, "step": 1825 }, { "entropy": 2.9258899450302125, "epoch": 0.5856, "grad_norm": 1.53125, "learning_rate": 0.00042040786527428335, "loss": 2.9065, "mean_token_accuracy": 0.44509958028793334, "num_tokens": 808374.0, "step": 1830 }, { "entropy": 2.959803581237793, "epoch": 0.5872, "grad_norm": 1.2109375, "learning_rate": 0.0004195068469647315, "loss": 2.903, "mean_token_accuracy": 0.44075829684734347, "num_tokens": 810632.0, "step": 1835 }, { "entropy": 2.9965363502502442, "epoch": 0.5888, "grad_norm": 1.1875, "learning_rate": 0.00041860186692697297, "loss": 2.9388, "mean_token_accuracy": 0.4424690276384354, "num_tokens": 812951.0, "step": 1840 }, { "entropy": 2.9641016483306886, "epoch": 0.5904, "grad_norm": 1.203125, "learning_rate": 0.00041769294997149264, "loss": 2.8931, "mean_token_accuracy": 0.44271588921546934, "num_tokens": 815168.0, "step": 1845 }, { "entropy": 2.9506234169006347, "epoch": 0.592, "grad_norm": 1.4140625, "learning_rate": 0.0004167801210167081, "loss": 2.9696, "mean_token_accuracy": 0.4374536544084549, "num_tokens": 817560.0, "step": 1850 }, { "entropy": 2.959603428840637, "epoch": 0.5936, "grad_norm": 1.15625, "learning_rate": 0.0004158634050882861, "loss": 2.9331, "mean_token_accuracy": 0.4436578154563904, "num_tokens": 819913.0, "step": 1855 }, { "entropy": 2.95600106716156, "epoch": 0.5952, "grad_norm": 1.6875, "learning_rate": 0.0004149428273184569, "loss": 2.9104, "mean_token_accuracy": 0.4492965519428253, "num_tokens": 822157.0, "step": 1860 }, { "entropy": 2.932568621635437, "epoch": 0.5968, "grad_norm": 1.1015625, "learning_rate": 0.0004140184129453253, "loss": 2.9006, "mean_token_accuracy": 0.44393897950649264, "num_tokens": 824319.0, "step": 1865 }, { "entropy": 2.842178702354431, "epoch": 0.5984, "grad_norm": 1.1171875, "learning_rate": 0.000413090187312178, "loss": 2.8294, "mean_token_accuracy": 0.46336430609226226, "num_tokens": 826336.0, "step": 1870 }, { "entropy": 2.932080864906311, "epoch": 0.6, "grad_norm": 1.1875, "learning_rate": 0.0004121581758667898, "loss": 2.9952, "mean_token_accuracy": 0.41981748342514036, "num_tokens": 828618.0, "step": 1875 }, { "entropy": 3.03024423122406, "epoch": 0.6016, "grad_norm": 2.125, "learning_rate": 0.00041122240416072533, "loss": 2.965, "mean_token_accuracy": 0.43029557168483734, "num_tokens": 830984.0, "step": 1880 }, { "entropy": 2.9398995637893677, "epoch": 0.6032, "grad_norm": 1.1796875, "learning_rate": 0.0004102828978486385, "loss": 2.9059, "mean_token_accuracy": 0.4471063494682312, "num_tokens": 833132.0, "step": 1885 }, { "entropy": 2.963112306594849, "epoch": 0.6048, "grad_norm": 1.3203125, "learning_rate": 0.0004093396826875695, "loss": 2.8954, "mean_token_accuracy": 0.4531849592924118, "num_tokens": 835222.0, "step": 1890 }, { "entropy": 2.997973084449768, "epoch": 0.6064, "grad_norm": 1.1796875, "learning_rate": 0.00040839278453623837, "loss": 2.9451, "mean_token_accuracy": 0.4340410023927689, "num_tokens": 837458.0, "step": 1895 }, { "entropy": 2.910131907463074, "epoch": 0.608, "grad_norm": 1.1015625, "learning_rate": 0.0004074422293543363, "loss": 2.8822, "mean_token_accuracy": 0.4447918087244034, "num_tokens": 839682.0, "step": 1900 }, { "entropy": 2.955132555961609, "epoch": 0.6096, "grad_norm": 1.203125, "learning_rate": 0.0004064880432018137, "loss": 2.9462, "mean_token_accuracy": 0.4323223292827606, "num_tokens": 841933.0, "step": 1905 }, { "entropy": 2.982530355453491, "epoch": 0.6112, "grad_norm": 1.7109375, "learning_rate": 0.00040553025223816615, "loss": 2.9529, "mean_token_accuracy": 0.443015918135643, "num_tokens": 844264.0, "step": 1910 }, { "entropy": 2.9218075275421143, "epoch": 0.6128, "grad_norm": 1.328125, "learning_rate": 0.00040456888272171653, "loss": 2.9212, "mean_token_accuracy": 0.44219004511833193, "num_tokens": 846514.0, "step": 1915 }, { "entropy": 2.975862002372742, "epoch": 0.6144, "grad_norm": 1.640625, "learning_rate": 0.00040360396100889577, "loss": 2.9567, "mean_token_accuracy": 0.4292555242776871, "num_tokens": 848723.0, "step": 1920 }, { "entropy": 2.963698649406433, "epoch": 0.616, "grad_norm": 1.109375, "learning_rate": 0.0004026355135535202, "loss": 2.878, "mean_token_accuracy": 0.4538561463356018, "num_tokens": 850908.0, "step": 1925 }, { "entropy": 2.9295140504837036, "epoch": 0.6176, "grad_norm": 1.0859375, "learning_rate": 0.000401663566906066, "loss": 2.9205, "mean_token_accuracy": 0.4405923217535019, "num_tokens": 853133.0, "step": 1930 }, { "entropy": 2.9326082468032837, "epoch": 0.6192, "grad_norm": 1.15625, "learning_rate": 0.00040068814771294134, "loss": 2.8567, "mean_token_accuracy": 0.453284427523613, "num_tokens": 855239.0, "step": 1935 }, { "entropy": 2.9947436094284057, "epoch": 0.6208, "grad_norm": 1.1796875, "learning_rate": 0.0003997092827157562, "loss": 2.9225, "mean_token_accuracy": 0.4441240161657333, "num_tokens": 857431.0, "step": 1940 }, { "entropy": 2.9254504442214966, "epoch": 0.6224, "grad_norm": 2.09375, "learning_rate": 0.000398726998750589, "loss": 2.8883, "mean_token_accuracy": 0.44197200536727904, "num_tokens": 859437.0, "step": 1945 }, { "entropy": 2.929787302017212, "epoch": 0.624, "grad_norm": 1.0625, "learning_rate": 0.00039774132274725076, "loss": 2.8881, "mean_token_accuracy": 0.4428322970867157, "num_tokens": 861510.0, "step": 1950 }, { "entropy": 3.0078527212142943, "epoch": 0.6256, "grad_norm": 1.21875, "learning_rate": 0.00039675228172854707, "loss": 2.9068, "mean_token_accuracy": 0.44754857420921323, "num_tokens": 863698.0, "step": 1955 }, { "entropy": 3.007625412940979, "epoch": 0.6272, "grad_norm": 1.3828125, "learning_rate": 0.0003957599028095371, "loss": 2.9049, "mean_token_accuracy": 0.4455776035785675, "num_tokens": 865923.0, "step": 1960 }, { "entropy": 2.957540273666382, "epoch": 0.6288, "grad_norm": 1.078125, "learning_rate": 0.00039476421319679017, "loss": 2.9675, "mean_token_accuracy": 0.4396782130002975, "num_tokens": 868247.0, "step": 1965 }, { "entropy": 2.942002534866333, "epoch": 0.6304, "grad_norm": 1.6328125, "learning_rate": 0.00039376524018764, "loss": 2.9008, "mean_token_accuracy": 0.4350299030542374, "num_tokens": 870557.0, "step": 1970 }, { "entropy": 2.902622389793396, "epoch": 0.632, "grad_norm": 0.99609375, "learning_rate": 0.00039276301116943616, "loss": 2.8809, "mean_token_accuracy": 0.4422562628984451, "num_tokens": 872742.0, "step": 1975 }, { "entropy": 2.9278737545013427, "epoch": 0.6336, "grad_norm": 1.1328125, "learning_rate": 0.0003917575536187936, "loss": 2.8592, "mean_token_accuracy": 0.4517789125442505, "num_tokens": 874814.0, "step": 1980 }, { "entropy": 2.940542984008789, "epoch": 0.6352, "grad_norm": 1.15625, "learning_rate": 0.00039074889510083894, "loss": 2.907, "mean_token_accuracy": 0.4450233310461044, "num_tokens": 876983.0, "step": 1985 }, { "entropy": 2.9536588191986084, "epoch": 0.6368, "grad_norm": 1.671875, "learning_rate": 0.00038973706326845495, "loss": 2.9322, "mean_token_accuracy": 0.4324341773986816, "num_tokens": 879259.0, "step": 1990 }, { "entropy": 2.990961289405823, "epoch": 0.6384, "grad_norm": 1.15625, "learning_rate": 0.0003887220858615225, "loss": 2.9238, "mean_token_accuracy": 0.44576185643672944, "num_tokens": 881526.0, "step": 1995 }, { "entropy": 2.872201347351074, "epoch": 0.64, "grad_norm": 1.28125, "learning_rate": 0.0003877039907061597, "loss": 2.8216, "mean_token_accuracy": 0.4564266115427017, "num_tokens": 883471.0, "step": 2000 }, { "entropy": 2.9151341676712037, "epoch": 0.6416, "grad_norm": 6.03125, "learning_rate": 0.0003866828057139598, "loss": 2.9031, "mean_token_accuracy": 0.4425373524427414, "num_tokens": 885706.0, "step": 2005 }, { "entropy": 2.9328678369522097, "epoch": 0.6432, "grad_norm": 1.15625, "learning_rate": 0.00038565855888122503, "loss": 2.872, "mean_token_accuracy": 0.447940868139267, "num_tokens": 887888.0, "step": 2010 }, { "entropy": 3.005712366104126, "epoch": 0.6448, "grad_norm": 1.625, "learning_rate": 0.00038463127828819975, "loss": 2.9234, "mean_token_accuracy": 0.4347133904695511, "num_tokens": 890182.0, "step": 2015 }, { "entropy": 2.9764514207839965, "epoch": 0.6464, "grad_norm": 1.546875, "learning_rate": 0.00038360099209830043, "loss": 2.9209, "mean_token_accuracy": 0.4349377155303955, "num_tokens": 892447.0, "step": 2020 }, { "entropy": 2.94652202129364, "epoch": 0.648, "grad_norm": 1.0390625, "learning_rate": 0.0003825677285573433, "loss": 2.9208, "mean_token_accuracy": 0.43728697299957275, "num_tokens": 894622.0, "step": 2025 }, { "entropy": 2.884352374076843, "epoch": 0.6496, "grad_norm": 1.1171875, "learning_rate": 0.00038153151599277027, "loss": 2.8508, "mean_token_accuracy": 0.45437749326229093, "num_tokens": 896694.0, "step": 2030 }, { "entropy": 2.883890223503113, "epoch": 0.6512, "grad_norm": 1.40625, "learning_rate": 0.0003804923828128723, "loss": 2.9306, "mean_token_accuracy": 0.44298937022686, "num_tokens": 898862.0, "step": 2035 }, { "entropy": 3.0047648668289186, "epoch": 0.6528, "grad_norm": 1.6484375, "learning_rate": 0.0003794503575060104, "loss": 2.9287, "mean_token_accuracy": 0.4394272893667221, "num_tokens": 901149.0, "step": 2040 }, { "entropy": 3.0060741662979127, "epoch": 0.6544, "grad_norm": 1.1484375, "learning_rate": 0.00037840546863983484, "loss": 2.9352, "mean_token_accuracy": 0.444516521692276, "num_tokens": 903477.0, "step": 2045 }, { "entropy": 2.985415482521057, "epoch": 0.656, "grad_norm": 1.3671875, "learning_rate": 0.0003773577448605015, "loss": 2.9641, "mean_token_accuracy": 0.42028396725654604, "num_tokens": 905702.0, "step": 2050 }, { "entropy": 3.01201274394989, "epoch": 0.6576, "grad_norm": 1.1171875, "learning_rate": 0.0003763072148918872, "loss": 2.9664, "mean_token_accuracy": 0.4346581995487213, "num_tokens": 908102.0, "step": 2055 }, { "entropy": 2.9599570751190187, "epoch": 0.6592, "grad_norm": 1.1328125, "learning_rate": 0.0003752539075348017, "loss": 2.9057, "mean_token_accuracy": 0.4458827257156372, "num_tokens": 910247.0, "step": 2060 }, { "entropy": 3.058842325210571, "epoch": 0.6608, "grad_norm": 1.2109375, "learning_rate": 0.00037419785166619817, "loss": 2.9694, "mean_token_accuracy": 0.4278321325778961, "num_tokens": 912536.0, "step": 2065 }, { "entropy": 3.0930434226989747, "epoch": 0.6624, "grad_norm": 1.09375, "learning_rate": 0.0003731390762383818, "loss": 2.8742, "mean_token_accuracy": 0.449863201379776, "num_tokens": 914653.0, "step": 2070 }, { "entropy": 2.9856122493743897, "epoch": 0.664, "grad_norm": 1.125, "learning_rate": 0.0003720776102782158, "loss": 2.9273, "mean_token_accuracy": 0.4387813091278076, "num_tokens": 916882.0, "step": 2075 }, { "entropy": 2.877395510673523, "epoch": 0.6656, "grad_norm": 1.1015625, "learning_rate": 0.00037101348288632555, "loss": 2.8471, "mean_token_accuracy": 0.46079154312610626, "num_tokens": 918922.0, "step": 2080 }, { "entropy": 2.970807933807373, "epoch": 0.6672, "grad_norm": 1.859375, "learning_rate": 0.0003699467232363012, "loss": 2.9398, "mean_token_accuracy": 0.43786206245422366, "num_tokens": 921072.0, "step": 2085 }, { "entropy": 3.0708029747009276, "epoch": 0.6688, "grad_norm": 1.0390625, "learning_rate": 0.0003688773605738973, "loss": 2.9533, "mean_token_accuracy": 0.4333233445882797, "num_tokens": 923523.0, "step": 2090 }, { "entropy": 3.0049412488937377, "epoch": 0.6704, "grad_norm": 1.046875, "learning_rate": 0.00036780542421623134, "loss": 2.9169, "mean_token_accuracy": 0.4365310609340668, "num_tokens": 925666.0, "step": 2095 }, { "entropy": 2.985308814048767, "epoch": 0.672, "grad_norm": 1.1015625, "learning_rate": 0.0003667309435509802, "loss": 2.9151, "mean_token_accuracy": 0.43432297110557555, "num_tokens": 927962.0, "step": 2100 }, { "entropy": 2.962909197807312, "epoch": 0.6736, "grad_norm": 1.2734375, "learning_rate": 0.0003656539480355741, "loss": 2.9056, "mean_token_accuracy": 0.44236772656440737, "num_tokens": 930138.0, "step": 2105 }, { "entropy": 2.885155940055847, "epoch": 0.6752, "grad_norm": 1.2265625, "learning_rate": 0.0003645744671963891, "loss": 2.8389, "mean_token_accuracy": 0.44500203132629396, "num_tokens": 932165.0, "step": 2110 }, { "entropy": 2.989500331878662, "epoch": 0.6768, "grad_norm": 1.109375, "learning_rate": 0.0003634925306279376, "loss": 2.8977, "mean_token_accuracy": 0.4393170565366745, "num_tokens": 934441.0, "step": 2115 }, { "entropy": 2.95395667552948, "epoch": 0.6784, "grad_norm": 0.97265625, "learning_rate": 0.0003624081679920574, "loss": 2.8932, "mean_token_accuracy": 0.4373865932226181, "num_tokens": 936564.0, "step": 2120 }, { "entropy": 2.9177491426467896, "epoch": 0.68, "grad_norm": 1.0859375, "learning_rate": 0.0003613214090170977, "loss": 2.8032, "mean_token_accuracy": 0.4614069700241089, "num_tokens": 938548.0, "step": 2125 }, { "entropy": 2.930582332611084, "epoch": 0.6816, "grad_norm": 1.3359375, "learning_rate": 0.0003602322834971048, "loss": 2.8357, "mean_token_accuracy": 0.44430472552776334, "num_tokens": 940679.0, "step": 2130 }, { "entropy": 2.988338589668274, "epoch": 0.6832, "grad_norm": 1.0625, "learning_rate": 0.0003591408212910051, "loss": 2.8328, "mean_token_accuracy": 0.4611793518066406, "num_tokens": 942686.0, "step": 2135 }, { "entropy": 2.9616580486297606, "epoch": 0.6848, "grad_norm": 1.4296875, "learning_rate": 0.0003580470523217863, "loss": 2.8884, "mean_token_accuracy": 0.456967830657959, "num_tokens": 944855.0, "step": 2140 }, { "entropy": 2.982660984992981, "epoch": 0.6864, "grad_norm": 1.1015625, "learning_rate": 0.0003569510065756771, "loss": 2.9079, "mean_token_accuracy": 0.4453257441520691, "num_tokens": 946999.0, "step": 2145 }, { "entropy": 3.002000093460083, "epoch": 0.688, "grad_norm": 1.828125, "learning_rate": 0.0003558527141013254, "loss": 2.8297, "mean_token_accuracy": 0.45135633647441864, "num_tokens": 949109.0, "step": 2150 }, { "entropy": 2.97209210395813, "epoch": 0.6896, "grad_norm": 1.1953125, "learning_rate": 0.0003547522050089742, "loss": 2.8852, "mean_token_accuracy": 0.44908978044986725, "num_tokens": 951293.0, "step": 2155 }, { "entropy": 2.933103632926941, "epoch": 0.6912, "grad_norm": 1.265625, "learning_rate": 0.00035364950946963606, "loss": 2.8488, "mean_token_accuracy": 0.4547416985034943, "num_tokens": 953466.0, "step": 2160 }, { "entropy": 3.0193636655807494, "epoch": 0.6928, "grad_norm": 1.1171875, "learning_rate": 0.0003525446577142663, "loss": 2.9363, "mean_token_accuracy": 0.4418988287448883, "num_tokens": 955827.0, "step": 2165 }, { "entropy": 2.973353314399719, "epoch": 0.6944, "grad_norm": 1.25, "learning_rate": 0.00035143768003293395, "loss": 2.8743, "mean_token_accuracy": 0.44199498295783995, "num_tokens": 957930.0, "step": 2170 }, { "entropy": 3.0493298530578614, "epoch": 0.696, "grad_norm": 1.1015625, "learning_rate": 0.0003503286067739913, "loss": 2.9361, "mean_token_accuracy": 0.43220597207546235, "num_tokens": 960257.0, "step": 2175 }, { "entropy": 3.0602797269821167, "epoch": 0.6976, "grad_norm": 1.59375, "learning_rate": 0.00034921746834324193, "loss": 2.9042, "mean_token_accuracy": 0.4409534424543381, "num_tokens": 962462.0, "step": 2180 }, { "entropy": 3.1046555519104, "epoch": 0.6992, "grad_norm": 1.3046875, "learning_rate": 0.0003481042952031072, "loss": 3.0035, "mean_token_accuracy": 0.42447165250778196, "num_tokens": 964961.0, "step": 2185 }, { "entropy": 2.9496065616607665, "epoch": 0.7008, "grad_norm": 1.484375, "learning_rate": 0.0003469891178717911, "loss": 2.8935, "mean_token_accuracy": 0.44119512140750883, "num_tokens": 967098.0, "step": 2190 }, { "entropy": 2.984723711013794, "epoch": 0.7024, "grad_norm": 1.3359375, "learning_rate": 0.0003458719669224436, "loss": 2.8743, "mean_token_accuracy": 0.4498720943927765, "num_tokens": 969320.0, "step": 2195 }, { "entropy": 3.009110188484192, "epoch": 0.704, "grad_norm": 1.171875, "learning_rate": 0.0003447528729823221, "loss": 2.9333, "mean_token_accuracy": 0.4297122836112976, "num_tokens": 971611.0, "step": 2200 }, { "entropy": 2.952143168449402, "epoch": 0.7056, "grad_norm": 1.40625, "learning_rate": 0.0003436318667319525, "loss": 2.8548, "mean_token_accuracy": 0.4506003618240356, "num_tokens": 973456.0, "step": 2205 }, { "entropy": 3.0895748376846313, "epoch": 0.7072, "grad_norm": 1.1328125, "learning_rate": 0.00034250897890428716, "loss": 2.8892, "mean_token_accuracy": 0.4512839734554291, "num_tokens": 975547.0, "step": 2210 }, { "entropy": 3.106333613395691, "epoch": 0.7088, "grad_norm": 1.46875, "learning_rate": 0.0003413842402838633, "loss": 2.9331, "mean_token_accuracy": 0.4400729775428772, "num_tokens": 977734.0, "step": 2215 }, { "entropy": 2.986264872550964, "epoch": 0.7104, "grad_norm": 1.1953125, "learning_rate": 0.00034025768170595834, "loss": 2.8363, "mean_token_accuracy": 0.44370077550411224, "num_tokens": 979832.0, "step": 2220 }, { "entropy": 3.00988655090332, "epoch": 0.712, "grad_norm": 1.3359375, "learning_rate": 0.0003391293340557446, "loss": 2.9068, "mean_token_accuracy": 0.4341572880744934, "num_tokens": 982023.0, "step": 2225 }, { "entropy": 3.0763514041900635, "epoch": 0.7136, "grad_norm": 1.5703125, "learning_rate": 0.0003379992282674431, "loss": 2.93, "mean_token_accuracy": 0.43178750574588776, "num_tokens": 984282.0, "step": 2230 }, { "entropy": 3.0266902446746826, "epoch": 0.7152, "grad_norm": 1.109375, "learning_rate": 0.0003368673953234749, "loss": 2.9098, "mean_token_accuracy": 0.43830609023571016, "num_tokens": 986521.0, "step": 2235 }, { "entropy": 3.0079190254211428, "epoch": 0.7168, "grad_norm": 1.3046875, "learning_rate": 0.00033573386625361176, "loss": 2.8589, "mean_token_accuracy": 0.4486089825630188, "num_tokens": 988646.0, "step": 2240 }, { "entropy": 3.0152955293655395, "epoch": 0.7184, "grad_norm": 1.0390625, "learning_rate": 0.00033459867213412567, "loss": 2.9259, "mean_token_accuracy": 0.43562940657138827, "num_tokens": 990968.0, "step": 2245 }, { "entropy": 2.977263402938843, "epoch": 0.72, "grad_norm": 1.1171875, "learning_rate": 0.000333461844086937, "loss": 2.8262, "mean_token_accuracy": 0.45436541438102723, "num_tokens": 992992.0, "step": 2250 }, { "entropy": 2.9837494611740114, "epoch": 0.7216, "grad_norm": 1.3671875, "learning_rate": 0.00033232341327876097, "loss": 2.8825, "mean_token_accuracy": 0.44874086380004885, "num_tokens": 995175.0, "step": 2255 }, { "entropy": 2.9914408206939695, "epoch": 0.7232, "grad_norm": 1.25, "learning_rate": 0.0003311834109202531, "loss": 2.9203, "mean_token_accuracy": 0.43564155995845794, "num_tokens": 997435.0, "step": 2260 }, { "entropy": 2.9815534114837647, "epoch": 0.7248, "grad_norm": 1.3125, "learning_rate": 0.00033004186826515416, "loss": 2.8832, "mean_token_accuracy": 0.4486095130443573, "num_tokens": 999481.0, "step": 2265 }, { "entropy": 3.0145650148391723, "epoch": 0.7264, "grad_norm": 1.234375, "learning_rate": 0.0003288988166094324, "loss": 2.9696, "mean_token_accuracy": 0.42953170239925387, "num_tokens": 1001954.0, "step": 2270 }, { "entropy": 2.9649643182754515, "epoch": 0.728, "grad_norm": 1.09375, "learning_rate": 0.00032775428729042656, "loss": 2.8459, "mean_token_accuracy": 0.4544994175434113, "num_tokens": 1004097.0, "step": 2275 }, { "entropy": 3.007507061958313, "epoch": 0.7296, "grad_norm": 0.96875, "learning_rate": 0.000326608311685986, "loss": 2.915, "mean_token_accuracy": 0.4395866394042969, "num_tokens": 1006390.0, "step": 2280 }, { "entropy": 3.086909604072571, "epoch": 0.7312, "grad_norm": 1.1015625, "learning_rate": 0.0003254609212136108, "loss": 2.9245, "mean_token_accuracy": 0.4338905721902847, "num_tokens": 1008767.0, "step": 2285 }, { "entropy": 3.0174469709396363, "epoch": 0.7328, "grad_norm": 1.09375, "learning_rate": 0.00032431214732959036, "loss": 2.918, "mean_token_accuracy": 0.4330415964126587, "num_tokens": 1010872.0, "step": 2290 }, { "entropy": 2.960281801223755, "epoch": 0.7344, "grad_norm": 1.3515625, "learning_rate": 0.000323162021528141, "loss": 2.8865, "mean_token_accuracy": 0.43942287266254426, "num_tokens": 1013092.0, "step": 2295 }, { "entropy": 2.9854196310043335, "epoch": 0.736, "grad_norm": 1.109375, "learning_rate": 0.00032201057534054264, "loss": 2.9163, "mean_token_accuracy": 0.43981327712535856, "num_tokens": 1015409.0, "step": 2300 }, { "entropy": 2.9264776706695557, "epoch": 0.7376, "grad_norm": 1.03125, "learning_rate": 0.00032085784033427414, "loss": 2.8929, "mean_token_accuracy": 0.44725373685359954, "num_tokens": 1017709.0, "step": 2305 }, { "entropy": 2.9120246887207033, "epoch": 0.7392, "grad_norm": 0.95703125, "learning_rate": 0.0003197038481121478, "loss": 2.8516, "mean_token_accuracy": 0.4545925796031952, "num_tokens": 1019798.0, "step": 2310 }, { "entropy": 2.8947165250778197, "epoch": 0.7408, "grad_norm": 1.0625, "learning_rate": 0.0003185486303114436, "loss": 2.8287, "mean_token_accuracy": 0.44671995341777804, "num_tokens": 1021830.0, "step": 2315 }, { "entropy": 2.947257924079895, "epoch": 0.7424, "grad_norm": 1.28125, "learning_rate": 0.0003173922186030409, "loss": 2.851, "mean_token_accuracy": 0.44511047899723055, "num_tokens": 1024019.0, "step": 2320 }, { "entropy": 2.9474929332733155, "epoch": 0.744, "grad_norm": 0.94140625, "learning_rate": 0.000316234644690551, "loss": 2.8927, "mean_token_accuracy": 0.4484694510698318, "num_tokens": 1026205.0, "step": 2325 }, { "entropy": 2.9977715969085694, "epoch": 0.7456, "grad_norm": 0.9296875, "learning_rate": 0.0003150759403094473, "loss": 2.9323, "mean_token_accuracy": 0.44033142030239103, "num_tokens": 1028581.0, "step": 2330 }, { "entropy": 2.964334225654602, "epoch": 0.7472, "grad_norm": 1.203125, "learning_rate": 0.00031391613722619587, "loss": 2.8387, "mean_token_accuracy": 0.4539773941040039, "num_tokens": 1030700.0, "step": 2335 }, { "entropy": 2.9664952993392943, "epoch": 0.7488, "grad_norm": 1.28125, "learning_rate": 0.000312755267237384, "loss": 2.9235, "mean_token_accuracy": 0.44700824916362764, "num_tokens": 1032967.0, "step": 2340 }, { "entropy": 2.9356072902679444, "epoch": 0.7504, "grad_norm": 1.140625, "learning_rate": 0.0003115933621688488, "loss": 2.8834, "mean_token_accuracy": 0.44536894261837007, "num_tokens": 1035201.0, "step": 2345 }, { "entropy": 2.931264066696167, "epoch": 0.752, "grad_norm": 1.1875, "learning_rate": 0.00031043045387480487, "loss": 2.8604, "mean_token_accuracy": 0.4562884718179703, "num_tokens": 1037433.0, "step": 2350 }, { "entropy": 2.8891395568847655, "epoch": 0.7536, "grad_norm": 1.1328125, "learning_rate": 0.0003092665742369703, "loss": 2.7734, "mean_token_accuracy": 0.465416020154953, "num_tokens": 1039350.0, "step": 2355 }, { "entropy": 2.8835833787918093, "epoch": 0.7552, "grad_norm": 1.4765625, "learning_rate": 0.00030810175516369343, "loss": 2.8739, "mean_token_accuracy": 0.4469373464584351, "num_tokens": 1041443.0, "step": 2360 }, { "entropy": 2.941758060455322, "epoch": 0.7568, "grad_norm": 1.1953125, "learning_rate": 0.0003069360285890775, "loss": 2.8719, "mean_token_accuracy": 0.44821923673152925, "num_tokens": 1043736.0, "step": 2365 }, { "entropy": 2.9532917976379394, "epoch": 0.7584, "grad_norm": 1.0859375, "learning_rate": 0.00030576942647210547, "loss": 2.884, "mean_token_accuracy": 0.4425459623336792, "num_tokens": 1045900.0, "step": 2370 }, { "entropy": 2.9466211795806885, "epoch": 0.76, "grad_norm": 1.3984375, "learning_rate": 0.00030460198079576355, "loss": 2.8945, "mean_token_accuracy": 0.4461941719055176, "num_tokens": 1048096.0, "step": 2375 }, { "entropy": 2.985423469543457, "epoch": 0.7616, "grad_norm": 0.9765625, "learning_rate": 0.0003034337235661648, "loss": 2.9139, "mean_token_accuracy": 0.44332720041275026, "num_tokens": 1050288.0, "step": 2380 }, { "entropy": 3.065624475479126, "epoch": 0.7632, "grad_norm": 1.2265625, "learning_rate": 0.0003022646868116714, "loss": 2.933, "mean_token_accuracy": 0.43668071627616883, "num_tokens": 1052709.0, "step": 2385 }, { "entropy": 3.0282908201217653, "epoch": 0.7648, "grad_norm": 1.21875, "learning_rate": 0.0003010949025820163, "loss": 2.9055, "mean_token_accuracy": 0.44049382507801055, "num_tokens": 1055008.0, "step": 2390 }, { "entropy": 2.971207594871521, "epoch": 0.7664, "grad_norm": 1.3359375, "learning_rate": 0.0002999244029474252, "loss": 2.8895, "mean_token_accuracy": 0.4490681529045105, "num_tokens": 1057268.0, "step": 2395 }, { "entropy": 2.8566708087921144, "epoch": 0.768, "grad_norm": 1.4375, "learning_rate": 0.00029875321999773684, "loss": 2.8463, "mean_token_accuracy": 0.4450335741043091, "num_tokens": 1059285.0, "step": 2400 }, { "entropy": 2.973609972000122, "epoch": 0.7696, "grad_norm": 1.8359375, "learning_rate": 0.00029758138584152333, "loss": 2.9612, "mean_token_accuracy": 0.42415072917938235, "num_tokens": 1061591.0, "step": 2405 }, { "entropy": 2.9553648710250853, "epoch": 0.7712, "grad_norm": 1.4296875, "learning_rate": 0.0002964089326052102, "loss": 2.8743, "mean_token_accuracy": 0.4511991322040558, "num_tokens": 1063736.0, "step": 2410 }, { "entropy": 2.984595799446106, "epoch": 0.7728, "grad_norm": 1.0234375, "learning_rate": 0.0002952358924321949, "loss": 2.9055, "mean_token_accuracy": 0.43462449312210083, "num_tokens": 1065953.0, "step": 2415 }, { "entropy": 3.0059932231903077, "epoch": 0.7744, "grad_norm": 1.0703125, "learning_rate": 0.00029406229748196657, "loss": 2.8948, "mean_token_accuracy": 0.44451610147953036, "num_tokens": 1068106.0, "step": 2420 }, { "entropy": 2.9257513523101806, "epoch": 0.776, "grad_norm": 1.1328125, "learning_rate": 0.0002928881799292235, "loss": 2.8144, "mean_token_accuracy": 0.46695643067359927, "num_tokens": 1070135.0, "step": 2425 }, { "entropy": 2.9223555088043214, "epoch": 0.7776, "grad_norm": 1.109375, "learning_rate": 0.00029171357196299154, "loss": 2.8557, "mean_token_accuracy": 0.45093517303466796, "num_tokens": 1072245.0, "step": 2430 }, { "entropy": 2.8693821668624877, "epoch": 0.7792, "grad_norm": 1.078125, "learning_rate": 0.0002905385057857414, "loss": 2.9073, "mean_token_accuracy": 0.4407109469175339, "num_tokens": 1074372.0, "step": 2435 }, { "entropy": 2.9372238874435426, "epoch": 0.7808, "grad_norm": 1.5234375, "learning_rate": 0.0002893630136125058, "loss": 2.878, "mean_token_accuracy": 0.44352578818798066, "num_tokens": 1076559.0, "step": 2440 }, { "entropy": 2.965376687049866, "epoch": 0.7824, "grad_norm": 1.171875, "learning_rate": 0.0002881871276699967, "loss": 2.8878, "mean_token_accuracy": 0.45097755789756777, "num_tokens": 1078724.0, "step": 2445 }, { "entropy": 2.950051736831665, "epoch": 0.784, "grad_norm": 1.3125, "learning_rate": 0.00028701088019572114, "loss": 2.8338, "mean_token_accuracy": 0.44691094756126404, "num_tokens": 1080869.0, "step": 2450 }, { "entropy": 2.961378741264343, "epoch": 0.7856, "grad_norm": 1.0390625, "learning_rate": 0.0002858343034370977, "loss": 2.8778, "mean_token_accuracy": 0.4402466356754303, "num_tokens": 1083017.0, "step": 2455 }, { "entropy": 2.951919937133789, "epoch": 0.7872, "grad_norm": 1.0078125, "learning_rate": 0.00028465742965057267, "loss": 2.8945, "mean_token_accuracy": 0.4406393527984619, "num_tokens": 1085173.0, "step": 2460 }, { "entropy": 2.9768404722213746, "epoch": 0.7888, "grad_norm": 0.92578125, "learning_rate": 0.00028348029110073533, "loss": 2.907, "mean_token_accuracy": 0.43652023673057555, "num_tokens": 1087437.0, "step": 2465 }, { "entropy": 2.955397605895996, "epoch": 0.7904, "grad_norm": 1.0625, "learning_rate": 0.00028230292005943365, "loss": 2.8638, "mean_token_accuracy": 0.4638360619544983, "num_tokens": 1089602.0, "step": 2470 }, { "entropy": 2.8937670469284056, "epoch": 0.792, "grad_norm": 1.390625, "learning_rate": 0.00028112534880488945, "loss": 2.8731, "mean_token_accuracy": 0.44972350001335143, "num_tokens": 1091673.0, "step": 2475 }, { "entropy": 2.9474738359451296, "epoch": 0.7936, "grad_norm": 0.98828125, "learning_rate": 0.0002799476096208137, "loss": 2.9004, "mean_token_accuracy": 0.4348028779029846, "num_tokens": 1093872.0, "step": 2480 }, { "entropy": 2.9170732736587524, "epoch": 0.7952, "grad_norm": 1.2890625, "learning_rate": 0.00027876973479552087, "loss": 2.8706, "mean_token_accuracy": 0.44423494637012484, "num_tokens": 1096076.0, "step": 2485 }, { "entropy": 3.031259369850159, "epoch": 0.7968, "grad_norm": 1.1015625, "learning_rate": 0.00027759175662104424, "loss": 2.9434, "mean_token_accuracy": 0.44129674732685087, "num_tokens": 1098528.0, "step": 2490 }, { "entropy": 2.9111172914505006, "epoch": 0.7984, "grad_norm": 1.1875, "learning_rate": 0.0002764137073922508, "loss": 2.7891, "mean_token_accuracy": 0.46722488701343534, "num_tokens": 1100575.0, "step": 2495 }, { "entropy": 2.9223501682281494, "epoch": 0.8, "grad_norm": 1.0625, "learning_rate": 0.00027523561940595505, "loss": 2.9217, "mean_token_accuracy": 0.4489959329366684, "num_tokens": 1102833.0, "step": 2500 }, { "entropy": 2.8666061639785765, "epoch": 0.8016, "grad_norm": 1.1484375, "learning_rate": 0.0002740575249600342, "loss": 2.9066, "mean_token_accuracy": 0.4415678709745407, "num_tokens": 1105110.0, "step": 2505 }, { "entropy": 2.920181465148926, "epoch": 0.8032, "grad_norm": 1.171875, "learning_rate": 0.00027287945635254263, "loss": 2.9202, "mean_token_accuracy": 0.4358911782503128, "num_tokens": 1107348.0, "step": 2510 }, { "entropy": 2.9587653160095213, "epoch": 0.8048, "grad_norm": 1.078125, "learning_rate": 0.00027170144588082635, "loss": 2.8697, "mean_token_accuracy": 0.44274553954601287, "num_tokens": 1109507.0, "step": 2515 }, { "entropy": 2.947556185722351, "epoch": 0.8064, "grad_norm": 1.109375, "learning_rate": 0.00027052352584063763, "loss": 2.8566, "mean_token_accuracy": 0.44831949174404145, "num_tokens": 1111502.0, "step": 2520 }, { "entropy": 3.010905075073242, "epoch": 0.808, "grad_norm": 1.015625, "learning_rate": 0.00026934572852524907, "loss": 2.9003, "mean_token_accuracy": 0.43905091285705566, "num_tokens": 1113746.0, "step": 2525 }, { "entropy": 2.947671890258789, "epoch": 0.8096, "grad_norm": 1.0234375, "learning_rate": 0.00026816808622456937, "loss": 2.8703, "mean_token_accuracy": 0.4387084603309631, "num_tokens": 1116030.0, "step": 2530 }, { "entropy": 2.883091449737549, "epoch": 0.8112, "grad_norm": 1.1328125, "learning_rate": 0.0002669906312242569, "loss": 2.8199, "mean_token_accuracy": 0.4524049639701843, "num_tokens": 1118102.0, "step": 2535 }, { "entropy": 2.8728092193603514, "epoch": 0.8128, "grad_norm": 1.0859375, "learning_rate": 0.00026581339580483525, "loss": 2.8828, "mean_token_accuracy": 0.4414022475481033, "num_tokens": 1120246.0, "step": 2540 }, { "entropy": 2.89880051612854, "epoch": 0.8144, "grad_norm": 1.0859375, "learning_rate": 0.0002646364122408082, "loss": 2.8368, "mean_token_accuracy": 0.45229865312576295, "num_tokens": 1122365.0, "step": 2545 }, { "entropy": 2.979207229614258, "epoch": 0.816, "grad_norm": 1.25, "learning_rate": 0.0002634597127997749, "loss": 2.9277, "mean_token_accuracy": 0.43487837016582487, "num_tokens": 1124651.0, "step": 2550 }, { "entropy": 2.9689237594604494, "epoch": 0.8176, "grad_norm": 1.171875, "learning_rate": 0.0002622833297415445, "loss": 2.8851, "mean_token_accuracy": 0.4453704297542572, "num_tokens": 1126826.0, "step": 2555 }, { "entropy": 2.912359046936035, "epoch": 0.8192, "grad_norm": 1.2265625, "learning_rate": 0.0002611072953172531, "loss": 2.8212, "mean_token_accuracy": 0.45373358726501467, "num_tokens": 1128925.0, "step": 2560 }, { "entropy": 2.9116891622543335, "epoch": 0.8208, "grad_norm": 1.453125, "learning_rate": 0.00025993164176847845, "loss": 2.94, "mean_token_accuracy": 0.4401213079690933, "num_tokens": 1131209.0, "step": 2565 }, { "entropy": 2.9820199728012087, "epoch": 0.8224, "grad_norm": 1.171875, "learning_rate": 0.0002587564013263564, "loss": 2.9732, "mean_token_accuracy": 0.4293658405542374, "num_tokens": 1133681.0, "step": 2570 }, { "entropy": 2.9426583528518675, "epoch": 0.824, "grad_norm": 1.1640625, "learning_rate": 0.0002575816062106974, "loss": 2.9092, "mean_token_accuracy": 0.4390734761953354, "num_tokens": 1135891.0, "step": 2575 }, { "entropy": 2.999843144416809, "epoch": 0.8256, "grad_norm": 1.1328125, "learning_rate": 0.00025640728862910293, "loss": 2.9403, "mean_token_accuracy": 0.43735610246658324, "num_tokens": 1138356.0, "step": 2580 }, { "entropy": 2.927291178703308, "epoch": 0.8272, "grad_norm": 1.0859375, "learning_rate": 0.00025523348077608285, "loss": 2.8654, "mean_token_accuracy": 0.4475594013929367, "num_tokens": 1140567.0, "step": 2585 }, { "entropy": 2.985684943199158, "epoch": 0.8288, "grad_norm": 1.4921875, "learning_rate": 0.00025406021483217225, "loss": 2.9121, "mean_token_accuracy": 0.4417661279439926, "num_tokens": 1142894.0, "step": 2590 }, { "entropy": 2.829986071586609, "epoch": 0.8304, "grad_norm": 1.7734375, "learning_rate": 0.00025288752296304963, "loss": 2.8038, "mean_token_accuracy": 0.4530776798725128, "num_tokens": 1144780.0, "step": 2595 }, { "entropy": 2.8750381231307984, "epoch": 0.832, "grad_norm": 1.078125, "learning_rate": 0.000251715437318655, "loss": 2.8211, "mean_token_accuracy": 0.46067106127738955, "num_tokens": 1146894.0, "step": 2600 }, { "entropy": 2.945680546760559, "epoch": 0.8336, "grad_norm": 1.140625, "learning_rate": 0.0002505439900323084, "loss": 2.9387, "mean_token_accuracy": 0.4324087291955948, "num_tokens": 1149168.0, "step": 2605 }, { "entropy": 2.8983019828796386, "epoch": 0.8352, "grad_norm": 1.515625, "learning_rate": 0.00024937321321982894, "loss": 2.8459, "mean_token_accuracy": 0.45197862982749937, "num_tokens": 1151320.0, "step": 2610 }, { "entropy": 2.894894599914551, "epoch": 0.8368, "grad_norm": 2.4375, "learning_rate": 0.00024820313897865433, "loss": 2.8957, "mean_token_accuracy": 0.44763479828834535, "num_tokens": 1153515.0, "step": 2615 }, { "entropy": 2.99886953830719, "epoch": 0.8384, "grad_norm": 1.09375, "learning_rate": 0.00024703379938696105, "loss": 2.9138, "mean_token_accuracy": 0.44238253831863406, "num_tokens": 1155859.0, "step": 2620 }, { "entropy": 2.9637449741363526, "epoch": 0.84, "grad_norm": 2.328125, "learning_rate": 0.00024586522650278447, "loss": 2.8946, "mean_token_accuracy": 0.44043884575366976, "num_tokens": 1158108.0, "step": 2625 }, { "entropy": 2.9379532814025877, "epoch": 0.8416, "grad_norm": 1.6328125, "learning_rate": 0.00024469745236314064, "loss": 2.8787, "mean_token_accuracy": 0.44664619863033295, "num_tokens": 1160258.0, "step": 2630 }, { "entropy": 2.924504518508911, "epoch": 0.8432, "grad_norm": 1.1640625, "learning_rate": 0.00024353050898314767, "loss": 2.9074, "mean_token_accuracy": 0.44169577956199646, "num_tokens": 1162388.0, "step": 2635 }, { "entropy": 2.915752816200256, "epoch": 0.8448, "grad_norm": 2.65625, "learning_rate": 0.00024236442835514743, "loss": 2.848, "mean_token_accuracy": 0.45955335199832914, "num_tokens": 1164495.0, "step": 2640 }, { "entropy": 2.9029451608657837, "epoch": 0.8464, "grad_norm": 1.4609375, "learning_rate": 0.00024119924244782965, "loss": 2.8846, "mean_token_accuracy": 0.44391528367996214, "num_tokens": 1166528.0, "step": 2645 }, { "entropy": 3.0527146577835085, "epoch": 0.848, "grad_norm": 1.2890625, "learning_rate": 0.00024003498320535462, "loss": 2.974, "mean_token_accuracy": 0.4264400273561478, "num_tokens": 1168868.0, "step": 2650 }, { "entropy": 3.060508465766907, "epoch": 0.8496, "grad_norm": 1.25, "learning_rate": 0.00023887168254647727, "loss": 2.9132, "mean_token_accuracy": 0.4397051393985748, "num_tokens": 1171066.0, "step": 2655 }, { "entropy": 2.967995810508728, "epoch": 0.8512, "grad_norm": 0.96484375, "learning_rate": 0.00023770937236367308, "loss": 2.9073, "mean_token_accuracy": 0.4405590325593948, "num_tokens": 1173416.0, "step": 2660 }, { "entropy": 2.9121148347854615, "epoch": 0.8528, "grad_norm": 1.2890625, "learning_rate": 0.00023654808452226278, "loss": 2.8591, "mean_token_accuracy": 0.4522093653678894, "num_tokens": 1175593.0, "step": 2665 }, { "entropy": 2.913937306404114, "epoch": 0.8544, "grad_norm": 1.1171875, "learning_rate": 0.00023538785085953912, "loss": 2.8452, "mean_token_accuracy": 0.457747283577919, "num_tokens": 1177681.0, "step": 2670 }, { "entropy": 2.881054997444153, "epoch": 0.856, "grad_norm": 1.1171875, "learning_rate": 0.00023422870318389404, "loss": 2.8688, "mean_token_accuracy": 0.44260228276252744, "num_tokens": 1179774.0, "step": 2675 }, { "entropy": 2.9047667741775514, "epoch": 0.8576, "grad_norm": 1.1875, "learning_rate": 0.0002330706732739468, "loss": 2.8582, "mean_token_accuracy": 0.4451305985450745, "num_tokens": 1181879.0, "step": 2680 }, { "entropy": 2.977963948249817, "epoch": 0.8592, "grad_norm": 1.1171875, "learning_rate": 0.00023191379287767211, "loss": 2.9565, "mean_token_accuracy": 0.4299389570951462, "num_tokens": 1184124.0, "step": 2685 }, { "entropy": 2.977361226081848, "epoch": 0.8608, "grad_norm": 1.5625, "learning_rate": 0.0002307580937115305, "loss": 2.9237, "mean_token_accuracy": 0.4347937345504761, "num_tokens": 1186474.0, "step": 2690 }, { "entropy": 3.003724551200867, "epoch": 0.8624, "grad_norm": 1.703125, "learning_rate": 0.00022960360745959846, "loss": 2.8684, "mean_token_accuracy": 0.4439554989337921, "num_tokens": 1188557.0, "step": 2695 }, { "entropy": 2.957149934768677, "epoch": 0.864, "grad_norm": 1.3046875, "learning_rate": 0.00022845036577269972, "loss": 2.8907, "mean_token_accuracy": 0.4393739640712738, "num_tokens": 1190733.0, "step": 2700 }, { "entropy": 2.9412027835845946, "epoch": 0.8656, "grad_norm": 1.359375, "learning_rate": 0.00022729840026753777, "loss": 2.867, "mean_token_accuracy": 0.4492296099662781, "num_tokens": 1192794.0, "step": 2705 }, { "entropy": 2.992423343658447, "epoch": 0.8672, "grad_norm": 1.2734375, "learning_rate": 0.0002261477425258287, "loss": 2.9147, "mean_token_accuracy": 0.44703419506549835, "num_tokens": 1195103.0, "step": 2710 }, { "entropy": 2.9774813890457152, "epoch": 0.8688, "grad_norm": 1.1171875, "learning_rate": 0.0002249984240934358, "loss": 2.9082, "mean_token_accuracy": 0.45050652623176574, "num_tokens": 1197307.0, "step": 2715 }, { "entropy": 2.9132537841796875, "epoch": 0.8704, "grad_norm": 1.078125, "learning_rate": 0.00022385047647950464, "loss": 2.8385, "mean_token_accuracy": 0.45678257942199707, "num_tokens": 1199360.0, "step": 2720 }, { "entropy": 2.9680255174636843, "epoch": 0.872, "grad_norm": 1.046875, "learning_rate": 0.0002227039311555986, "loss": 2.9003, "mean_token_accuracy": 0.44852447509765625, "num_tokens": 1201638.0, "step": 2725 }, { "entropy": 2.9566770792007446, "epoch": 0.8736, "grad_norm": 1.2421875, "learning_rate": 0.0002215588195548372, "loss": 2.9018, "mean_token_accuracy": 0.4464226752519608, "num_tokens": 1203963.0, "step": 2730 }, { "entropy": 2.9229081153869627, "epoch": 0.8752, "grad_norm": 0.9609375, "learning_rate": 0.00022041517307103337, "loss": 2.8697, "mean_token_accuracy": 0.45343998074531555, "num_tokens": 1206185.0, "step": 2735 }, { "entropy": 2.9380676984786986, "epoch": 0.8768, "grad_norm": 1.234375, "learning_rate": 0.0002192730230578331, "loss": 2.8441, "mean_token_accuracy": 0.4449337601661682, "num_tokens": 1208351.0, "step": 2740 }, { "entropy": 2.961194062232971, "epoch": 0.8784, "grad_norm": 1.015625, "learning_rate": 0.0002181324008278559, "loss": 2.8848, "mean_token_accuracy": 0.44320608079433443, "num_tokens": 1210604.0, "step": 2745 }, { "entropy": 2.984656810760498, "epoch": 0.88, "grad_norm": 2.53125, "learning_rate": 0.00021699333765183655, "loss": 2.8955, "mean_token_accuracy": 0.4382534384727478, "num_tokens": 1212884.0, "step": 2750 }, { "entropy": 2.922535252571106, "epoch": 0.8816, "grad_norm": 1.109375, "learning_rate": 0.0002158558647577673, "loss": 2.8609, "mean_token_accuracy": 0.452529639005661, "num_tokens": 1214988.0, "step": 2755 }, { "entropy": 2.921734118461609, "epoch": 0.8832, "grad_norm": 1.0234375, "learning_rate": 0.00021472001333004215, "loss": 2.8568, "mean_token_accuracy": 0.4529965102672577, "num_tokens": 1217294.0, "step": 2760 }, { "entropy": 2.9577472448349, "epoch": 0.8848, "grad_norm": 1.0234375, "learning_rate": 0.00021358581450860186, "loss": 2.844, "mean_token_accuracy": 0.45522329211235046, "num_tokens": 1219595.0, "step": 2765 }, { "entropy": 2.897872066497803, "epoch": 0.8864, "grad_norm": 1.2109375, "learning_rate": 0.0002124532993880799, "loss": 2.8422, "mean_token_accuracy": 0.4452970206737518, "num_tokens": 1221764.0, "step": 2770 }, { "entropy": 2.8961577892303465, "epoch": 0.888, "grad_norm": 0.9921875, "learning_rate": 0.00021132249901695044, "loss": 2.8746, "mean_token_accuracy": 0.4422593981027603, "num_tokens": 1224093.0, "step": 2775 }, { "entropy": 2.894803762435913, "epoch": 0.8896, "grad_norm": 1.875, "learning_rate": 0.00021019344439667705, "loss": 2.882, "mean_token_accuracy": 0.4526560217142105, "num_tokens": 1226289.0, "step": 2780 }, { "entropy": 3.0008044242858887, "epoch": 0.8912, "grad_norm": 1.0234375, "learning_rate": 0.00020906616648086213, "loss": 2.9416, "mean_token_accuracy": 0.4286386549472809, "num_tokens": 1228740.0, "step": 2785 }, { "entropy": 2.9415072917938234, "epoch": 0.8928, "grad_norm": 1.09375, "learning_rate": 0.00020794069617439942, "loss": 2.8724, "mean_token_accuracy": 0.43999955654144285, "num_tokens": 1230963.0, "step": 2790 }, { "entropy": 2.9294766902923586, "epoch": 0.8944, "grad_norm": 1.1015625, "learning_rate": 0.00020681706433262593, "loss": 2.8356, "mean_token_accuracy": 0.44667740166187286, "num_tokens": 1232995.0, "step": 2795 }, { "entropy": 2.968204212188721, "epoch": 0.896, "grad_norm": 1.1640625, "learning_rate": 0.00020569530176047602, "loss": 2.9386, "mean_token_accuracy": 0.43440439403057096, "num_tokens": 1235384.0, "step": 2800 }, { "entropy": 2.9155136823654173, "epoch": 0.8976, "grad_norm": 1.234375, "learning_rate": 0.0002045754392116374, "loss": 2.8238, "mean_token_accuracy": 0.4479335069656372, "num_tokens": 1237417.0, "step": 2805 }, { "entropy": 2.954969549179077, "epoch": 0.8992, "grad_norm": 1.1640625, "learning_rate": 0.00020345750738770757, "loss": 2.8416, "mean_token_accuracy": 0.4532474637031555, "num_tokens": 1239539.0, "step": 2810 }, { "entropy": 2.943054366111755, "epoch": 0.9008, "grad_norm": 1.1484375, "learning_rate": 0.00020234153693735214, "loss": 2.8456, "mean_token_accuracy": 0.4507363855838776, "num_tokens": 1241762.0, "step": 2815 }, { "entropy": 2.9151533365249636, "epoch": 0.9024, "grad_norm": 0.96875, "learning_rate": 0.0002012275584554647, "loss": 2.8615, "mean_token_accuracy": 0.4435191124677658, "num_tokens": 1243848.0, "step": 2820 }, { "entropy": 2.964799904823303, "epoch": 0.904, "grad_norm": 1.1171875, "learning_rate": 0.00020011560248232803, "loss": 2.8829, "mean_token_accuracy": 0.4484334021806717, "num_tokens": 1246041.0, "step": 2825 }, { "entropy": 2.9201528072357177, "epoch": 0.9056, "grad_norm": 1.0625, "learning_rate": 0.00019900569950277692, "loss": 2.8239, "mean_token_accuracy": 0.45506506264209745, "num_tokens": 1248172.0, "step": 2830 }, { "entropy": 2.9409337043762207, "epoch": 0.9072, "grad_norm": 0.9140625, "learning_rate": 0.00019789787994536228, "loss": 2.8994, "mean_token_accuracy": 0.43933017551898956, "num_tokens": 1250479.0, "step": 2835 }, { "entropy": 2.922648048400879, "epoch": 0.9088, "grad_norm": 0.9609375, "learning_rate": 0.00019679217418151667, "loss": 2.8778, "mean_token_accuracy": 0.44817441403865815, "num_tokens": 1252701.0, "step": 2840 }, { "entropy": 2.914081406593323, "epoch": 0.9104, "grad_norm": 1.2109375, "learning_rate": 0.00019568861252472236, "loss": 2.8645, "mean_token_accuracy": 0.448718985915184, "num_tokens": 1254805.0, "step": 2845 }, { "entropy": 2.940555787086487, "epoch": 0.912, "grad_norm": 1.1953125, "learning_rate": 0.00019458722522967952, "loss": 2.9154, "mean_token_accuracy": 0.440675950050354, "num_tokens": 1257116.0, "step": 2850 }, { "entropy": 3.0115319013595583, "epoch": 0.9136, "grad_norm": 1.046875, "learning_rate": 0.00019348804249147723, "loss": 2.9453, "mean_token_accuracy": 0.4348461925983429, "num_tokens": 1259591.0, "step": 2855 }, { "entropy": 3.0097128629684446, "epoch": 0.9152, "grad_norm": 1.5234375, "learning_rate": 0.0001923910944447655, "loss": 2.9047, "mean_token_accuracy": 0.44291664361953736, "num_tokens": 1261843.0, "step": 2860 }, { "entropy": 2.957931566238403, "epoch": 0.9168, "grad_norm": 0.9296875, "learning_rate": 0.00019129641116292928, "loss": 2.8786, "mean_token_accuracy": 0.4521709203720093, "num_tokens": 1264048.0, "step": 2865 }, { "entropy": 2.963532829284668, "epoch": 0.9184, "grad_norm": 1.0, "learning_rate": 0.00019020402265726343, "loss": 2.8946, "mean_token_accuracy": 0.4415064722299576, "num_tokens": 1266278.0, "step": 2870 }, { "entropy": 3.008980083465576, "epoch": 0.92, "grad_norm": 1.0, "learning_rate": 0.0001891139588761509, "loss": 2.9334, "mean_token_accuracy": 0.4405760496854782, "num_tokens": 1268672.0, "step": 2875 }, { "entropy": 3.0282814264297486, "epoch": 0.9216, "grad_norm": 1.1953125, "learning_rate": 0.00018802624970424076, "loss": 2.9627, "mean_token_accuracy": 0.43039882481098174, "num_tokens": 1271131.0, "step": 2880 }, { "entropy": 2.977272367477417, "epoch": 0.9232, "grad_norm": 1.0625, "learning_rate": 0.00018694092496162945, "loss": 2.9092, "mean_token_accuracy": 0.43852547705173495, "num_tokens": 1273355.0, "step": 2885 }, { "entropy": 2.9832343816757203, "epoch": 0.9248, "grad_norm": 1.03125, "learning_rate": 0.00018585801440304306, "loss": 2.933, "mean_token_accuracy": 0.4351432114839554, "num_tokens": 1275673.0, "step": 2890 }, { "entropy": 2.932812213897705, "epoch": 0.9264, "grad_norm": 1.0078125, "learning_rate": 0.00018477754771702165, "loss": 2.8326, "mean_token_accuracy": 0.45522419214248655, "num_tokens": 1277773.0, "step": 2895 }, { "entropy": 3.003762054443359, "epoch": 0.928, "grad_norm": 1.6328125, "learning_rate": 0.00018369955452510506, "loss": 2.9153, "mean_token_accuracy": 0.4390197813510895, "num_tokens": 1280091.0, "step": 2900 }, { "entropy": 2.911535930633545, "epoch": 0.9296, "grad_norm": 1.046875, "learning_rate": 0.0001826240643810212, "loss": 2.8358, "mean_token_accuracy": 0.4533025860786438, "num_tokens": 1282186.0, "step": 2905 }, { "entropy": 2.9808855295181274, "epoch": 0.9312, "grad_norm": 1.2734375, "learning_rate": 0.0001815511067698758, "loss": 2.9145, "mean_token_accuracy": 0.44003509283065795, "num_tokens": 1284549.0, "step": 2910 }, { "entropy": 2.9279879093170167, "epoch": 0.9328, "grad_norm": 0.984375, "learning_rate": 0.0001804807111073436, "loss": 2.8986, "mean_token_accuracy": 0.44652416110038756, "num_tokens": 1286727.0, "step": 2915 }, { "entropy": 2.9614415645599363, "epoch": 0.9344, "grad_norm": 1.03125, "learning_rate": 0.0001794129067388625, "loss": 2.8973, "mean_token_accuracy": 0.43919517695903776, "num_tokens": 1288879.0, "step": 2920 }, { "entropy": 2.8971622705459597, "epoch": 0.936, "grad_norm": 1.078125, "learning_rate": 0.00017834772293882868, "loss": 2.815, "mean_token_accuracy": 0.45427608788013457, "num_tokens": 1290944.0, "step": 2925 }, { "entropy": 2.983414721488953, "epoch": 0.9376, "grad_norm": 1.1015625, "learning_rate": 0.000177285188909794, "loss": 2.9166, "mean_token_accuracy": 0.4338977485895157, "num_tokens": 1293319.0, "step": 2930 }, { "entropy": 2.9010852813720702, "epoch": 0.9392, "grad_norm": 1.3984375, "learning_rate": 0.0001762253337816656, "loss": 2.8131, "mean_token_accuracy": 0.45710650384426116, "num_tokens": 1295298.0, "step": 2935 }, { "entropy": 2.912653946876526, "epoch": 0.9408, "grad_norm": 1.1328125, "learning_rate": 0.00017516818661090738, "loss": 2.8508, "mean_token_accuracy": 0.458192965388298, "num_tokens": 1297600.0, "step": 2940 }, { "entropy": 2.9122406005859376, "epoch": 0.9424, "grad_norm": 0.90625, "learning_rate": 0.0001741137763797428, "loss": 2.8918, "mean_token_accuracy": 0.44095310270786287, "num_tokens": 1299824.0, "step": 2945 }, { "entropy": 2.923125743865967, "epoch": 0.944, "grad_norm": 1.078125, "learning_rate": 0.00017306213199536115, "loss": 2.9127, "mean_token_accuracy": 0.4380089223384857, "num_tokens": 1302097.0, "step": 2950 }, { "entropy": 2.9138781309127806, "epoch": 0.9456, "grad_norm": 1.1484375, "learning_rate": 0.0001720132822891243, "loss": 2.8426, "mean_token_accuracy": 0.4544492304325104, "num_tokens": 1304279.0, "step": 2955 }, { "entropy": 2.8925965309143065, "epoch": 0.9472, "grad_norm": 0.9921875, "learning_rate": 0.0001709672560157769, "loss": 2.8919, "mean_token_accuracy": 0.4376676261425018, "num_tokens": 1306566.0, "step": 2960 }, { "entropy": 2.9171841382980346, "epoch": 0.9488, "grad_norm": 1.1015625, "learning_rate": 0.00016992408185265758, "loss": 2.8431, "mean_token_accuracy": 0.4547513484954834, "num_tokens": 1308632.0, "step": 2965 }, { "entropy": 2.9132872104644774, "epoch": 0.9504, "grad_norm": 1.140625, "learning_rate": 0.00016888378839891298, "loss": 2.9194, "mean_token_accuracy": 0.43245779275894164, "num_tokens": 1310920.0, "step": 2970 }, { "entropy": 2.92850923538208, "epoch": 0.952, "grad_norm": 1.21875, "learning_rate": 0.0001678464041747137, "loss": 2.8856, "mean_token_accuracy": 0.44850285947322843, "num_tokens": 1313149.0, "step": 2975 }, { "entropy": 2.858417201042175, "epoch": 0.9536, "grad_norm": 1.125, "learning_rate": 0.00016681195762047223, "loss": 2.7997, "mean_token_accuracy": 0.45988663136959074, "num_tokens": 1315167.0, "step": 2980 }, { "entropy": 2.96560423374176, "epoch": 0.9552, "grad_norm": 1.0703125, "learning_rate": 0.00016578047709606337, "loss": 2.9403, "mean_token_accuracy": 0.4384672850370407, "num_tokens": 1317554.0, "step": 2985 }, { "entropy": 2.936877202987671, "epoch": 0.9568, "grad_norm": 1.1171875, "learning_rate": 0.00016475199088004678, "loss": 2.8865, "mean_token_accuracy": 0.4436915129423141, "num_tokens": 1319839.0, "step": 2990 }, { "entropy": 2.889366102218628, "epoch": 0.9584, "grad_norm": 1.1953125, "learning_rate": 0.00016372652716889163, "loss": 2.8318, "mean_token_accuracy": 0.4491864800453186, "num_tokens": 1321830.0, "step": 2995 }, { "entropy": 2.9192376375198363, "epoch": 0.96, "grad_norm": 1.0234375, "learning_rate": 0.0001627041140762035, "loss": 2.8774, "mean_token_accuracy": 0.448089799284935, "num_tokens": 1324013.0, "step": 3000 }, { "entropy": 2.822802686691284, "epoch": 0.9616, "grad_norm": 1.078125, "learning_rate": 0.00016168477963195382, "loss": 2.7919, "mean_token_accuracy": 0.46399557292461396, "num_tokens": 1325968.0, "step": 3005 }, { "entropy": 2.902009677886963, "epoch": 0.9632, "grad_norm": 1.3203125, "learning_rate": 0.0001606685517817114, "loss": 2.9027, "mean_token_accuracy": 0.42942934632301333, "num_tokens": 1328223.0, "step": 3010 }, { "entropy": 2.969492554664612, "epoch": 0.9648, "grad_norm": 1.0390625, "learning_rate": 0.00015965545838587592, "loss": 2.8935, "mean_token_accuracy": 0.4412230223417282, "num_tokens": 1330619.0, "step": 3015 }, { "entropy": 2.93003191947937, "epoch": 0.9664, "grad_norm": 1.109375, "learning_rate": 0.00015864552721891467, "loss": 2.8825, "mean_token_accuracy": 0.4463993698358536, "num_tokens": 1332770.0, "step": 3020 }, { "entropy": 2.935991358757019, "epoch": 0.968, "grad_norm": 1.1015625, "learning_rate": 0.00015763878596860076, "loss": 2.8649, "mean_token_accuracy": 0.4543608367443085, "num_tokens": 1335057.0, "step": 3025 }, { "entropy": 2.9606276512145997, "epoch": 0.9696, "grad_norm": 1.1640625, "learning_rate": 0.00015663526223525412, "loss": 2.8927, "mean_token_accuracy": 0.43624573945999146, "num_tokens": 1337238.0, "step": 3030 }, { "entropy": 2.9343493223190307, "epoch": 0.9712, "grad_norm": 0.95703125, "learning_rate": 0.0001556349835309848, "loss": 2.8662, "mean_token_accuracy": 0.4551668256521225, "num_tokens": 1339447.0, "step": 3035 }, { "entropy": 2.9501862049102785, "epoch": 0.9728, "grad_norm": 1.1953125, "learning_rate": 0.0001546379772789389, "loss": 2.9101, "mean_token_accuracy": 0.4377577006816864, "num_tokens": 1341849.0, "step": 3040 }, { "entropy": 2.8754868507385254, "epoch": 0.9744, "grad_norm": 0.9609375, "learning_rate": 0.00015364427081254622, "loss": 2.8459, "mean_token_accuracy": 0.45830716490745543, "num_tokens": 1344079.0, "step": 3045 }, { "entropy": 2.8969892978668215, "epoch": 0.976, "grad_norm": 0.91015625, "learning_rate": 0.00015265389137477165, "loss": 2.8535, "mean_token_accuracy": 0.4491910576820374, "num_tokens": 1346191.0, "step": 3050 }, { "entropy": 2.937625861167908, "epoch": 0.9776, "grad_norm": 1.0703125, "learning_rate": 0.00015166686611736786, "loss": 2.8547, "mean_token_accuracy": 0.44432043135166166, "num_tokens": 1348523.0, "step": 3055 }, { "entropy": 2.998405599594116, "epoch": 0.9792, "grad_norm": 0.9765625, "learning_rate": 0.00015068322210013064, "loss": 2.9304, "mean_token_accuracy": 0.4362600237131119, "num_tokens": 1350869.0, "step": 3060 }, { "entropy": 2.9206026554107667, "epoch": 0.9808, "grad_norm": 0.93359375, "learning_rate": 0.0001497029862901578, "loss": 2.8919, "mean_token_accuracy": 0.44179760813713076, "num_tokens": 1353137.0, "step": 3065 }, { "entropy": 2.8943419456481934, "epoch": 0.9824, "grad_norm": 1.1875, "learning_rate": 0.00014872618556110905, "loss": 2.8588, "mean_token_accuracy": 0.4497217029333115, "num_tokens": 1355383.0, "step": 3070 }, { "entropy": 2.836391234397888, "epoch": 0.984, "grad_norm": 1.53125, "learning_rate": 0.00014775284669246992, "loss": 2.8225, "mean_token_accuracy": 0.4554732948541641, "num_tokens": 1357429.0, "step": 3075 }, { "entropy": 2.8723347902297975, "epoch": 0.9856, "grad_norm": 1.1328125, "learning_rate": 0.00014678299636881716, "loss": 2.8539, "mean_token_accuracy": 0.44728320837020874, "num_tokens": 1359560.0, "step": 3080 }, { "entropy": 2.93394558429718, "epoch": 0.9872, "grad_norm": 0.94921875, "learning_rate": 0.0001458166611790873, "loss": 2.8676, "mean_token_accuracy": 0.44904575049877166, "num_tokens": 1361928.0, "step": 3085 }, { "entropy": 2.8841625690460204, "epoch": 0.9888, "grad_norm": 1.328125, "learning_rate": 0.00014485386761584773, "loss": 2.8271, "mean_token_accuracy": 0.4506414353847504, "num_tokens": 1364078.0, "step": 3090 }, { "entropy": 2.93695387840271, "epoch": 0.9904, "grad_norm": 1.0234375, "learning_rate": 0.00014389464207457042, "loss": 2.8872, "mean_token_accuracy": 0.44658905267715454, "num_tokens": 1366361.0, "step": 3095 }, { "entropy": 2.880967140197754, "epoch": 0.992, "grad_norm": 1.203125, "learning_rate": 0.00014293901085290795, "loss": 2.8645, "mean_token_accuracy": 0.4420303195714951, "num_tokens": 1368518.0, "step": 3100 }, { "entropy": 2.8800176858901976, "epoch": 0.9936, "grad_norm": 1.09375, "learning_rate": 0.00014198700014997307, "loss": 2.8379, "mean_token_accuracy": 0.4591920435428619, "num_tokens": 1370691.0, "step": 3105 }, { "entropy": 2.875255084037781, "epoch": 0.9952, "grad_norm": 1.2734375, "learning_rate": 0.00014103863606562016, "loss": 2.8831, "mean_token_accuracy": 0.43948179185390474, "num_tokens": 1372830.0, "step": 3110 }, { "entropy": 2.9139184951782227, "epoch": 0.9968, "grad_norm": 1.03125, "learning_rate": 0.00014009394459972964, "loss": 2.8683, "mean_token_accuracy": 0.442516627907753, "num_tokens": 1374985.0, "step": 3115 }, { "entropy": 2.8382085800170898, "epoch": 0.9984, "grad_norm": 1.0625, "learning_rate": 0.00013915295165149513, "loss": 2.7851, "mean_token_accuracy": 0.4656050056219101, "num_tokens": 1376908.0, "step": 3120 }, { "entropy": 2.850643539428711, "epoch": 1.0, "grad_norm": 1.03125, "learning_rate": 0.00013821568301871384, "loss": 2.8018, "mean_token_accuracy": 0.4667865127325058, "num_tokens": 1378880.0, "step": 3125 }, { "entropy": 2.910154700279236, "epoch": 1.0016, "grad_norm": 0.98828125, "learning_rate": 0.00013728216439707862, "loss": 2.875, "mean_token_accuracy": 0.4438807636499405, "num_tokens": 1381159.0, "step": 3130 }, { "entropy": 2.905166411399841, "epoch": 1.0032, "grad_norm": 1.1796875, "learning_rate": 0.00013635242137947419, "loss": 2.8758, "mean_token_accuracy": 0.4487916111946106, "num_tokens": 1383337.0, "step": 3135 }, { "entropy": 2.918097805976868, "epoch": 1.0048, "grad_norm": 1.171875, "learning_rate": 0.00013542647945527498, "loss": 2.9228, "mean_token_accuracy": 0.43111771941184995, "num_tokens": 1385646.0, "step": 3140 }, { "entropy": 2.9552776336669924, "epoch": 1.0064, "grad_norm": 1.1796875, "learning_rate": 0.0001345043640096465, "loss": 2.8883, "mean_token_accuracy": 0.4475600093603134, "num_tokens": 1387904.0, "step": 3145 }, { "entropy": 2.9810845851898193, "epoch": 1.008, "grad_norm": 1.140625, "learning_rate": 0.00013358610032284957, "loss": 2.8984, "mean_token_accuracy": 0.44070603549480436, "num_tokens": 1390172.0, "step": 3150 }, { "entropy": 2.953131651878357, "epoch": 1.0096, "grad_norm": 0.98828125, "learning_rate": 0.000132671713569547, "loss": 2.8389, "mean_token_accuracy": 0.4531745284795761, "num_tokens": 1392320.0, "step": 3155 }, { "entropy": 2.9414817094802856, "epoch": 1.0112, "grad_norm": 1.0859375, "learning_rate": 0.0001317612288181136, "loss": 2.8234, "mean_token_accuracy": 0.4625133275985718, "num_tokens": 1394445.0, "step": 3160 }, { "entropy": 2.9300568103790283, "epoch": 1.0128, "grad_norm": 1.328125, "learning_rate": 0.00013085467102994864, "loss": 2.8177, "mean_token_accuracy": 0.45458551347255705, "num_tokens": 1396633.0, "step": 3165 }, { "entropy": 2.8585967779159547, "epoch": 1.0144, "grad_norm": 1.015625, "learning_rate": 0.00012995206505879198, "loss": 2.8253, "mean_token_accuracy": 0.45586331784725187, "num_tokens": 1398673.0, "step": 3170 }, { "entropy": 2.8420963525772094, "epoch": 1.016, "grad_norm": 1.015625, "learning_rate": 0.0001290534356500421, "loss": 2.8197, "mean_token_accuracy": 0.45721601247787474, "num_tokens": 1400698.0, "step": 3175 }, { "entropy": 2.840663743019104, "epoch": 1.0176, "grad_norm": 1.1328125, "learning_rate": 0.00012815880744007827, "loss": 2.8747, "mean_token_accuracy": 0.44926016628742216, "num_tokens": 1402832.0, "step": 3180 }, { "entropy": 2.866150140762329, "epoch": 1.0192, "grad_norm": 1.03125, "learning_rate": 0.00012726820495558483, "loss": 2.8413, "mean_token_accuracy": 0.44752993881702424, "num_tokens": 1404982.0, "step": 3185 }, { "entropy": 2.892368483543396, "epoch": 1.0208, "grad_norm": 1.0703125, "learning_rate": 0.00012638165261287868, "loss": 2.8654, "mean_token_accuracy": 0.44305120408535004, "num_tokens": 1407080.0, "step": 3190 }, { "entropy": 2.9561034202575684, "epoch": 1.0224, "grad_norm": 1.0625, "learning_rate": 0.0001254991747172402, "loss": 2.9097, "mean_token_accuracy": 0.44477667212486266, "num_tokens": 1409320.0, "step": 3195 }, { "entropy": 2.991245460510254, "epoch": 1.024, "grad_norm": 1.015625, "learning_rate": 0.00012462079546224662, "loss": 2.9044, "mean_token_accuracy": 0.4459333002567291, "num_tokens": 1411464.0, "step": 3200 }, { "entropy": 2.9974965810775758, "epoch": 1.0256, "grad_norm": 1.296875, "learning_rate": 0.00012374653892910896, "loss": 2.8468, "mean_token_accuracy": 0.4494890421628952, "num_tokens": 1413667.0, "step": 3205 }, { "entropy": 2.943142366409302, "epoch": 1.0272, "grad_norm": 1.015625, "learning_rate": 0.00012287642908601166, "loss": 2.8447, "mean_token_accuracy": 0.45564105808734895, "num_tokens": 1415778.0, "step": 3210 }, { "entropy": 2.960755467414856, "epoch": 1.0288, "grad_norm": 1.484375, "learning_rate": 0.00012201048978745569, "loss": 2.8806, "mean_token_accuracy": 0.44263093173503876, "num_tokens": 1417964.0, "step": 3215 }, { "entropy": 2.952089858055115, "epoch": 1.0304, "grad_norm": 1.0703125, "learning_rate": 0.00012114874477360427, "loss": 2.9424, "mean_token_accuracy": 0.4364638030529022, "num_tokens": 1420246.0, "step": 3220 }, { "entropy": 2.9189977645874023, "epoch": 1.032, "grad_norm": 1.21875, "learning_rate": 0.00012029121766963236, "loss": 2.854, "mean_token_accuracy": 0.45222486555576324, "num_tokens": 1422427.0, "step": 3225 }, { "entropy": 2.925573539733887, "epoch": 1.0336, "grad_norm": 1.015625, "learning_rate": 0.00011943793198507858, "loss": 2.844, "mean_token_accuracy": 0.45238340795040133, "num_tokens": 1424701.0, "step": 3230 }, { "entropy": 2.8957201719284056, "epoch": 1.0352, "grad_norm": 0.96875, "learning_rate": 0.00011858891111320104, "loss": 2.8552, "mean_token_accuracy": 0.4456094205379486, "num_tokens": 1426916.0, "step": 3235 }, { "entropy": 2.877080202102661, "epoch": 1.0368, "grad_norm": 1.0625, "learning_rate": 0.0001177441783303359, "loss": 2.8028, "mean_token_accuracy": 0.4573970854282379, "num_tokens": 1428940.0, "step": 3240 }, { "entropy": 2.8774147748947145, "epoch": 1.0384, "grad_norm": 1.125, "learning_rate": 0.00011690375679525896, "loss": 2.8132, "mean_token_accuracy": 0.4620885670185089, "num_tokens": 1430949.0, "step": 3245 }, { "entropy": 2.916015863418579, "epoch": 1.04, "grad_norm": 1.0703125, "learning_rate": 0.00011606766954855124, "loss": 2.8356, "mean_token_accuracy": 0.4443682163953781, "num_tokens": 1433067.0, "step": 3250 }, { "entropy": 2.894275665283203, "epoch": 1.0416, "grad_norm": 1.125, "learning_rate": 0.00011523593951196702, "loss": 2.8064, "mean_token_accuracy": 0.46011479794979093, "num_tokens": 1435074.0, "step": 3255 }, { "entropy": 2.8826929092407227, "epoch": 1.0432, "grad_norm": 1.0703125, "learning_rate": 0.00011440858948780523, "loss": 2.8284, "mean_token_accuracy": 0.45082894861698153, "num_tokens": 1437186.0, "step": 3260 }, { "entropy": 2.8625028133392334, "epoch": 1.0448, "grad_norm": 1.03125, "learning_rate": 0.00011358564215828484, "loss": 2.8016, "mean_token_accuracy": 0.45917061865329745, "num_tokens": 1439270.0, "step": 3265 }, { "entropy": 2.919316792488098, "epoch": 1.0464, "grad_norm": 1.015625, "learning_rate": 0.00011276712008492254, "loss": 2.8822, "mean_token_accuracy": 0.4417962789535522, "num_tokens": 1441391.0, "step": 3270 }, { "entropy": 2.9069796562194825, "epoch": 1.048, "grad_norm": 0.9921875, "learning_rate": 0.00011195304570791451, "loss": 2.8697, "mean_token_accuracy": 0.45134317576885224, "num_tokens": 1443615.0, "step": 3275 }, { "entropy": 2.9291406631469727, "epoch": 1.0496, "grad_norm": 1.2890625, "learning_rate": 0.00011114344134552094, "loss": 2.8579, "mean_token_accuracy": 0.4531704008579254, "num_tokens": 1445796.0, "step": 3280 }, { "entropy": 2.948936033248901, "epoch": 1.0512, "grad_norm": 1.0390625, "learning_rate": 0.0001103383291934545, "loss": 2.8329, "mean_token_accuracy": 0.45455129742622374, "num_tokens": 1448115.0, "step": 3285 }, { "entropy": 2.9454702138900757, "epoch": 1.0528, "grad_norm": 1.578125, "learning_rate": 0.00010953773132427141, "loss": 2.846, "mean_token_accuracy": 0.45498157739639283, "num_tokens": 1450286.0, "step": 3290 }, { "entropy": 2.922731876373291, "epoch": 1.0544, "grad_norm": 1.078125, "learning_rate": 0.00010874166968676677, "loss": 2.8767, "mean_token_accuracy": 0.44795788526535035, "num_tokens": 1452437.0, "step": 3295 }, { "entropy": 2.9160258531570435, "epoch": 1.056, "grad_norm": 1.90625, "learning_rate": 0.00010795016610537251, "loss": 2.8819, "mean_token_accuracy": 0.4423990875482559, "num_tokens": 1454718.0, "step": 3300 }, { "entropy": 2.969084930419922, "epoch": 1.0576, "grad_norm": 0.984375, "learning_rate": 0.00010716324227955904, "loss": 2.899, "mean_token_accuracy": 0.4472406208515167, "num_tokens": 1457131.0, "step": 3305 }, { "entropy": 2.9933769702911377, "epoch": 1.0592, "grad_norm": 1.0, "learning_rate": 0.0001063809197832406, "loss": 2.9151, "mean_token_accuracy": 0.43931878805160524, "num_tokens": 1459692.0, "step": 3310 }, { "entropy": 2.9689890861511232, "epoch": 1.0608, "grad_norm": 1.1015625, "learning_rate": 0.00010560322006418368, "loss": 2.9316, "mean_token_accuracy": 0.44082845449447633, "num_tokens": 1462140.0, "step": 3315 }, { "entropy": 2.937236261367798, "epoch": 1.0624, "grad_norm": 1.0, "learning_rate": 0.00010483016444341887, "loss": 2.8928, "mean_token_accuracy": 0.44165705144405365, "num_tokens": 1464408.0, "step": 3320 }, { "entropy": 2.8921984910964964, "epoch": 1.064, "grad_norm": 1.109375, "learning_rate": 0.00010406177411465654, "loss": 2.8613, "mean_token_accuracy": 0.45629408955574036, "num_tokens": 1466636.0, "step": 3325 }, { "entropy": 2.917197847366333, "epoch": 1.0656, "grad_norm": 0.93359375, "learning_rate": 0.00010329807014370562, "loss": 2.8695, "mean_token_accuracy": 0.458002632856369, "num_tokens": 1468854.0, "step": 3330 }, { "entropy": 2.937503457069397, "epoch": 1.0672, "grad_norm": 1.015625, "learning_rate": 0.00010253907346789632, "loss": 2.8217, "mean_token_accuracy": 0.45761874318122864, "num_tokens": 1471015.0, "step": 3335 }, { "entropy": 2.86787006855011, "epoch": 1.0688, "grad_norm": 1.015625, "learning_rate": 0.00010178480489550596, "loss": 2.8362, "mean_token_accuracy": 0.4549447178840637, "num_tokens": 1473100.0, "step": 3340 }, { "entropy": 2.92385847568512, "epoch": 1.0704, "grad_norm": 1.15625, "learning_rate": 0.00010103528510518836, "loss": 2.9116, "mean_token_accuracy": 0.43709237277507784, "num_tokens": 1475346.0, "step": 3345 }, { "entropy": 2.934638023376465, "epoch": 1.072, "grad_norm": 1.0390625, "learning_rate": 0.0001002905346454073, "loss": 2.8609, "mean_token_accuracy": 0.45075803399086, "num_tokens": 1477710.0, "step": 3350 }, { "entropy": 2.9242029190063477, "epoch": 1.0735999999999999, "grad_norm": 1.2578125, "learning_rate": 9.955057393387285e-05, "loss": 2.9334, "mean_token_accuracy": 0.43468915224075316, "num_tokens": 1480049.0, "step": 3355 }, { "entropy": 2.899898362159729, "epoch": 1.0752, "grad_norm": 1.03125, "learning_rate": 9.88154232569816e-05, "loss": 2.8693, "mean_token_accuracy": 0.4400544613599777, "num_tokens": 1482190.0, "step": 3360 }, { "entropy": 2.9233818292617797, "epoch": 1.0768, "grad_norm": 1.1484375, "learning_rate": 9.808510276926075e-05, "loss": 2.8393, "mean_token_accuracy": 0.45188432931900024, "num_tokens": 1484326.0, "step": 3365 }, { "entropy": 2.9737884044647216, "epoch": 1.0784, "grad_norm": 1.15625, "learning_rate": 9.735963249281549e-05, "loss": 2.864, "mean_token_accuracy": 0.44917926490306853, "num_tokens": 1486599.0, "step": 3370 }, { "entropy": 2.9760022878646852, "epoch": 1.08, "grad_norm": 1.09375, "learning_rate": 9.663903231677974e-05, "loss": 2.8464, "mean_token_accuracy": 0.45147477090358734, "num_tokens": 1488905.0, "step": 3375 }, { "entropy": 2.908150005340576, "epoch": 1.0816, "grad_norm": 0.8671875, "learning_rate": 9.592332199677145e-05, "loss": 2.8532, "mean_token_accuracy": 0.45783216059207915, "num_tokens": 1491171.0, "step": 3380 }, { "entropy": 2.935186982154846, "epoch": 1.0832, "grad_norm": 1.3125, "learning_rate": 9.521252115435061e-05, "loss": 2.9299, "mean_token_accuracy": 0.4383124619722366, "num_tokens": 1493513.0, "step": 3385 }, { "entropy": 2.8872111320495604, "epoch": 1.0848, "grad_norm": 1.25, "learning_rate": 9.450664927648126e-05, "loss": 2.8446, "mean_token_accuracy": 0.4421195387840271, "num_tokens": 1495668.0, "step": 3390 }, { "entropy": 2.8951022148132326, "epoch": 1.0864, "grad_norm": 0.9921875, "learning_rate": 9.380572571499758e-05, "loss": 2.8645, "mean_token_accuracy": 0.44183502793312074, "num_tokens": 1497792.0, "step": 3395 }, { "entropy": 2.916491222381592, "epoch": 1.088, "grad_norm": 1.0234375, "learning_rate": 9.310976968607307e-05, "loss": 2.9204, "mean_token_accuracy": 0.43581128120422363, "num_tokens": 1500074.0, "step": 3400 }, { "entropy": 2.884718489646912, "epoch": 1.0896, "grad_norm": 1.1796875, "learning_rate": 9.241880026969381e-05, "loss": 2.8427, "mean_token_accuracy": 0.4484182119369507, "num_tokens": 1502133.0, "step": 3405 }, { "entropy": 2.9245787143707274, "epoch": 1.0912, "grad_norm": 1.1796875, "learning_rate": 9.173283640913537e-05, "loss": 2.8829, "mean_token_accuracy": 0.45576434433460233, "num_tokens": 1504316.0, "step": 3410 }, { "entropy": 2.9601083755493165, "epoch": 1.0928, "grad_norm": 1.0625, "learning_rate": 9.10518969104436e-05, "loss": 2.8951, "mean_token_accuracy": 0.44321806728839874, "num_tokens": 1506639.0, "step": 3415 }, { "entropy": 2.9153964042663576, "epoch": 1.0944, "grad_norm": 1.0703125, "learning_rate": 9.037600044191868e-05, "loss": 2.8734, "mean_token_accuracy": 0.4405162364244461, "num_tokens": 1508863.0, "step": 3420 }, { "entropy": 2.9321192502975464, "epoch": 1.096, "grad_norm": 0.91796875, "learning_rate": 8.970516553360383e-05, "loss": 2.8522, "mean_token_accuracy": 0.4530280500650406, "num_tokens": 1511119.0, "step": 3425 }, { "entropy": 2.973708438873291, "epoch": 1.0976, "grad_norm": 1.078125, "learning_rate": 8.903941057677692e-05, "loss": 2.866, "mean_token_accuracy": 0.4498842030763626, "num_tokens": 1513380.0, "step": 3430 }, { "entropy": 2.9102531909942626, "epoch": 1.0992, "grad_norm": 1.1171875, "learning_rate": 8.837875382344635e-05, "loss": 2.8607, "mean_token_accuracy": 0.4471791386604309, "num_tokens": 1515454.0, "step": 3435 }, { "entropy": 2.939423656463623, "epoch": 1.1008, "grad_norm": 1.1640625, "learning_rate": 8.772321338585076e-05, "loss": 2.9195, "mean_token_accuracy": 0.4250449001789093, "num_tokens": 1517788.0, "step": 3440 }, { "entropy": 2.9743568181991575, "epoch": 1.1024, "grad_norm": 1.0234375, "learning_rate": 8.707280723596242e-05, "loss": 2.895, "mean_token_accuracy": 0.4467761397361755, "num_tokens": 1520120.0, "step": 3445 }, { "entropy": 2.9497989654541015, "epoch": 1.104, "grad_norm": 1.171875, "learning_rate": 8.64275532049944e-05, "loss": 2.9049, "mean_token_accuracy": 0.44069608449935915, "num_tokens": 1522414.0, "step": 3450 }, { "entropy": 2.9272958993911744, "epoch": 1.1056, "grad_norm": 1.2734375, "learning_rate": 8.578746898291198e-05, "loss": 2.8858, "mean_token_accuracy": 0.43844324052333833, "num_tokens": 1524494.0, "step": 3455 }, { "entropy": 2.8901722192764283, "epoch": 1.1072, "grad_norm": 0.8828125, "learning_rate": 8.515257211794742e-05, "loss": 2.8374, "mean_token_accuracy": 0.4545303225517273, "num_tokens": 1526730.0, "step": 3460 }, { "entropy": 2.9210543870925902, "epoch": 1.1088, "grad_norm": 1.0234375, "learning_rate": 8.452288001611896e-05, "loss": 2.8584, "mean_token_accuracy": 0.44453992545604704, "num_tokens": 1528903.0, "step": 3465 }, { "entropy": 2.9284739017486574, "epoch": 1.1104, "grad_norm": 1.171875, "learning_rate": 8.389840994075379e-05, "loss": 2.8728, "mean_token_accuracy": 0.44965373873710635, "num_tokens": 1531066.0, "step": 3470 }, { "entropy": 2.974892830848694, "epoch": 1.112, "grad_norm": 1.234375, "learning_rate": 8.327917901201435e-05, "loss": 2.9215, "mean_token_accuracy": 0.44402619302272794, "num_tokens": 1533410.0, "step": 3475 }, { "entropy": 2.8764452934265137, "epoch": 1.1136, "grad_norm": 1.2890625, "learning_rate": 8.266520420642931e-05, "loss": 2.902, "mean_token_accuracy": 0.4433110415935516, "num_tokens": 1535637.0, "step": 3480 }, { "entropy": 2.9302066564559937, "epoch": 1.1152, "grad_norm": 1.4296875, "learning_rate": 8.205650235642828e-05, "loss": 2.8948, "mean_token_accuracy": 0.44368058145046235, "num_tokens": 1537899.0, "step": 3485 }, { "entropy": 2.9187855243682863, "epoch": 1.1168, "grad_norm": 1.1171875, "learning_rate": 8.145309014987978e-05, "loss": 2.8599, "mean_token_accuracy": 0.44634734094142914, "num_tokens": 1540044.0, "step": 3490 }, { "entropy": 2.9369451284408568, "epoch": 1.1184, "grad_norm": 1.1875, "learning_rate": 8.085498412963437e-05, "loss": 2.8574, "mean_token_accuracy": 0.4445057541131973, "num_tokens": 1542253.0, "step": 3495 }, { "entropy": 2.882553219795227, "epoch": 1.12, "grad_norm": 1.0, "learning_rate": 8.026220069307078e-05, "loss": 2.8034, "mean_token_accuracy": 0.4587402492761612, "num_tokens": 1544313.0, "step": 3500 }, { "entropy": 2.9640145778656004, "epoch": 1.1216, "grad_norm": 1.1328125, "learning_rate": 7.967475609164621e-05, "loss": 2.8935, "mean_token_accuracy": 0.4417268455028534, "num_tokens": 1546517.0, "step": 3505 }, { "entropy": 2.9522467136383055, "epoch": 1.1232, "grad_norm": 1.6328125, "learning_rate": 7.909266643045124e-05, "loss": 2.8619, "mean_token_accuracy": 0.44433446824550626, "num_tokens": 1548795.0, "step": 3510 }, { "entropy": 2.9092732667922974, "epoch": 1.1248, "grad_norm": 1.28125, "learning_rate": 7.851594766776802e-05, "loss": 2.8506, "mean_token_accuracy": 0.4580541640520096, "num_tokens": 1550915.0, "step": 3515 }, { "entropy": 2.90768780708313, "epoch": 1.1264, "grad_norm": 1.078125, "learning_rate": 7.794461561463265e-05, "loss": 2.8293, "mean_token_accuracy": 0.4552948325872421, "num_tokens": 1552996.0, "step": 3520 }, { "entropy": 2.9287519454956055, "epoch": 1.1280000000000001, "grad_norm": 1.0703125, "learning_rate": 7.7378685934402e-05, "loss": 2.8842, "mean_token_accuracy": 0.44655353128910064, "num_tokens": 1555198.0, "step": 3525 }, { "entropy": 2.9286380052566527, "epoch": 1.1296, "grad_norm": 1.125, "learning_rate": 7.68181741423242e-05, "loss": 2.8378, "mean_token_accuracy": 0.4591426819562912, "num_tokens": 1557421.0, "step": 3530 }, { "entropy": 2.8866475343704225, "epoch": 1.1312, "grad_norm": 1.046875, "learning_rate": 7.626309560511313e-05, "loss": 2.7942, "mean_token_accuracy": 0.46680617332458496, "num_tokens": 1559513.0, "step": 3535 }, { "entropy": 3.023671770095825, "epoch": 1.1328, "grad_norm": 0.91796875, "learning_rate": 7.571346554052724e-05, "loss": 2.9671, "mean_token_accuracy": 0.4291446566581726, "num_tokens": 1562035.0, "step": 3540 }, { "entropy": 2.921654486656189, "epoch": 1.1344, "grad_norm": 1.0546875, "learning_rate": 7.516929901695249e-05, "loss": 2.8644, "mean_token_accuracy": 0.4626066416501999, "num_tokens": 1564319.0, "step": 3545 }, { "entropy": 2.979964852333069, "epoch": 1.1360000000000001, "grad_norm": 0.9609375, "learning_rate": 7.463061095298893e-05, "loss": 2.8956, "mean_token_accuracy": 0.4382535606622696, "num_tokens": 1566591.0, "step": 3550 }, { "entropy": 2.9657538414001463, "epoch": 1.1376, "grad_norm": 0.92578125, "learning_rate": 7.409741611704198e-05, "loss": 2.9334, "mean_token_accuracy": 0.4299424201250076, "num_tokens": 1568915.0, "step": 3555 }, { "entropy": 2.9413408756256105, "epoch": 1.1392, "grad_norm": 1.1328125, "learning_rate": 7.35697291269174e-05, "loss": 2.8925, "mean_token_accuracy": 0.43368988335132597, "num_tokens": 1571133.0, "step": 3560 }, { "entropy": 2.9115732192993162, "epoch": 1.1408, "grad_norm": 1.0, "learning_rate": 7.304756444942056e-05, "loss": 2.8694, "mean_token_accuracy": 0.45736863315105436, "num_tokens": 1573271.0, "step": 3565 }, { "entropy": 2.9249136447906494, "epoch": 1.1424, "grad_norm": 1.1171875, "learning_rate": 7.253093639995994e-05, "loss": 2.8605, "mean_token_accuracy": 0.4405524283647537, "num_tokens": 1575520.0, "step": 3570 }, { "entropy": 2.924656796455383, "epoch": 1.144, "grad_norm": 0.9921875, "learning_rate": 7.20198591421544e-05, "loss": 2.8773, "mean_token_accuracy": 0.4417353868484497, "num_tokens": 1577648.0, "step": 3575 }, { "entropy": 2.9047645807266234, "epoch": 1.1456, "grad_norm": 1.171875, "learning_rate": 7.151434668744517e-05, "loss": 2.836, "mean_token_accuracy": 0.4575980305671692, "num_tokens": 1579846.0, "step": 3580 }, { "entropy": 2.9089220762252808, "epoch": 1.1472, "grad_norm": 1.078125, "learning_rate": 7.101441289471153e-05, "loss": 2.8336, "mean_token_accuracy": 0.4530055969953537, "num_tokens": 1581987.0, "step": 3585 }, { "entropy": 2.9274566650390623, "epoch": 1.1488, "grad_norm": 1.0625, "learning_rate": 7.052007146989098e-05, "loss": 2.8302, "mean_token_accuracy": 0.46225800216197965, "num_tokens": 1584114.0, "step": 3590 }, { "entropy": 2.9081597089767457, "epoch": 1.1504, "grad_norm": 0.9765625, "learning_rate": 7.003133596560341e-05, "loss": 2.8534, "mean_token_accuracy": 0.4496993154287338, "num_tokens": 1586316.0, "step": 3595 }, { "entropy": 2.9959617614746095, "epoch": 1.152, "grad_norm": 0.95703125, "learning_rate": 6.954821978077952e-05, "loss": 2.924, "mean_token_accuracy": 0.4415579229593277, "num_tokens": 1588827.0, "step": 3600 }, { "entropy": 2.9230748414993286, "epoch": 1.1536, "grad_norm": 1.0, "learning_rate": 6.907073616029356e-05, "loss": 2.8597, "mean_token_accuracy": 0.4494464576244354, "num_tokens": 1591037.0, "step": 3605 }, { "entropy": 2.954933786392212, "epoch": 1.1552, "grad_norm": 1.0703125, "learning_rate": 6.85988981946002e-05, "loss": 2.9154, "mean_token_accuracy": 0.44115144312381743, "num_tokens": 1593398.0, "step": 3610 }, { "entropy": 2.918868064880371, "epoch": 1.1568, "grad_norm": 1.453125, "learning_rate": 6.813271881937564e-05, "loss": 2.8817, "mean_token_accuracy": 0.4488807737827301, "num_tokens": 1595758.0, "step": 3615 }, { "entropy": 2.9787248373031616, "epoch": 1.1584, "grad_norm": 1.046875, "learning_rate": 6.767221081516286e-05, "loss": 2.9022, "mean_token_accuracy": 0.4427231132984161, "num_tokens": 1598130.0, "step": 3620 }, { "entropy": 2.9282839775085447, "epoch": 1.16, "grad_norm": 1.28125, "learning_rate": 6.72173868070215e-05, "loss": 2.8389, "mean_token_accuracy": 0.45102006793022154, "num_tokens": 1600288.0, "step": 3625 }, { "entropy": 2.9464649200439452, "epoch": 1.1616, "grad_norm": 1.09375, "learning_rate": 6.676825926418149e-05, "loss": 2.8324, "mean_token_accuracy": 0.4627915441989899, "num_tokens": 1602463.0, "step": 3630 }, { "entropy": 2.946268105506897, "epoch": 1.1632, "grad_norm": 1.1015625, "learning_rate": 6.632484049970122e-05, "loss": 2.9155, "mean_token_accuracy": 0.43858610689640043, "num_tokens": 1604802.0, "step": 3635 }, { "entropy": 2.9082115411758425, "epoch": 1.1648, "grad_norm": 1.03125, "learning_rate": 6.588714267013019e-05, "loss": 2.8611, "mean_token_accuracy": 0.44852556586265563, "num_tokens": 1606837.0, "step": 3640 }, { "entropy": 2.9096720218658447, "epoch": 1.1663999999999999, "grad_norm": 1.0546875, "learning_rate": 6.545517777517544e-05, "loss": 2.8639, "mean_token_accuracy": 0.45403674244880676, "num_tokens": 1609092.0, "step": 3645 }, { "entropy": 2.906810092926025, "epoch": 1.168, "grad_norm": 1.0546875, "learning_rate": 6.502895765737281e-05, "loss": 2.835, "mean_token_accuracy": 0.4538911312818527, "num_tokens": 1611264.0, "step": 3650 }, { "entropy": 2.9424187183380126, "epoch": 1.1696, "grad_norm": 1.1484375, "learning_rate": 6.460849400176212e-05, "loss": 2.8892, "mean_token_accuracy": 0.4430447190999985, "num_tokens": 1613555.0, "step": 3655 }, { "entropy": 2.9268627166748047, "epoch": 1.1712, "grad_norm": 0.91796875, "learning_rate": 6.419379833556694e-05, "loss": 2.8331, "mean_token_accuracy": 0.4631327152252197, "num_tokens": 1615810.0, "step": 3660 }, { "entropy": 2.8886401414871217, "epoch": 1.1728, "grad_norm": 2.46875, "learning_rate": 6.378488202787835e-05, "loss": 2.8791, "mean_token_accuracy": 0.4445572793483734, "num_tokens": 1617907.0, "step": 3665 }, { "entropy": 2.9272811889648436, "epoch": 1.1743999999999999, "grad_norm": 1.1484375, "learning_rate": 6.33817562893435e-05, "loss": 2.8843, "mean_token_accuracy": 0.44224337339401243, "num_tokens": 1620191.0, "step": 3670 }, { "entropy": 2.900171995162964, "epoch": 1.176, "grad_norm": 0.9296875, "learning_rate": 6.29844321718582e-05, "loss": 2.824, "mean_token_accuracy": 0.45361705422401427, "num_tokens": 1622352.0, "step": 3675 }, { "entropy": 2.9959622621536255, "epoch": 1.1776, "grad_norm": 1.0234375, "learning_rate": 6.259292056826383e-05, "loss": 2.894, "mean_token_accuracy": 0.4437588781118393, "num_tokens": 1624684.0, "step": 3680 }, { "entropy": 2.936618137359619, "epoch": 1.1792, "grad_norm": 1.078125, "learning_rate": 6.220723221204873e-05, "loss": 2.9108, "mean_token_accuracy": 0.44253872334957123, "num_tokens": 1627035.0, "step": 3685 }, { "entropy": 2.948612189292908, "epoch": 1.1808, "grad_norm": 1.0390625, "learning_rate": 6.182737767705406e-05, "loss": 2.8718, "mean_token_accuracy": 0.4431992769241333, "num_tokens": 1629356.0, "step": 3690 }, { "entropy": 2.9471986293792725, "epoch": 1.1824, "grad_norm": 1.0859375, "learning_rate": 6.145336737718375e-05, "loss": 2.8737, "mean_token_accuracy": 0.44046359360218046, "num_tokens": 1631478.0, "step": 3695 }, { "entropy": 2.9031818389892576, "epoch": 1.184, "grad_norm": 1.03125, "learning_rate": 6.10852115661191e-05, "loss": 2.8417, "mean_token_accuracy": 0.4519147574901581, "num_tokens": 1633588.0, "step": 3700 }, { "entropy": 2.9622613430023192, "epoch": 1.1856, "grad_norm": 1.1328125, "learning_rate": 6.072292033703766e-05, "loss": 2.8771, "mean_token_accuracy": 0.45182370245456693, "num_tokens": 1635940.0, "step": 3705 }, { "entropy": 2.882323884963989, "epoch": 1.1872, "grad_norm": 1.1484375, "learning_rate": 6.036650362233648e-05, "loss": 2.8389, "mean_token_accuracy": 0.45641362071037295, "num_tokens": 1638066.0, "step": 3710 }, { "entropy": 2.891980743408203, "epoch": 1.1888, "grad_norm": 1.078125, "learning_rate": 6.0015971193359824e-05, "loss": 2.855, "mean_token_accuracy": 0.4433140456676483, "num_tokens": 1640266.0, "step": 3715 }, { "entropy": 2.922886872291565, "epoch": 1.1904, "grad_norm": 1.0078125, "learning_rate": 5.9671332660131306e-05, "loss": 2.865, "mean_token_accuracy": 0.4502765327692032, "num_tokens": 1642410.0, "step": 3720 }, { "entropy": 2.9298436403274537, "epoch": 1.192, "grad_norm": 1.015625, "learning_rate": 5.933259747109042e-05, "loss": 2.8302, "mean_token_accuracy": 0.45269047021865844, "num_tokens": 1644547.0, "step": 3725 }, { "entropy": 2.888952207565308, "epoch": 1.1936, "grad_norm": 1.0859375, "learning_rate": 5.899977491283351e-05, "loss": 2.8456, "mean_token_accuracy": 0.4499317049980164, "num_tokens": 1646646.0, "step": 3730 }, { "entropy": 2.956196689605713, "epoch": 1.1952, "grad_norm": 1.0390625, "learning_rate": 5.867287410985908e-05, "loss": 2.8871, "mean_token_accuracy": 0.44261796176433565, "num_tokens": 1648882.0, "step": 3735 }, { "entropy": 2.9073150396347045, "epoch": 1.1968, "grad_norm": 1.0078125, "learning_rate": 5.835190402431779e-05, "loss": 2.8533, "mean_token_accuracy": 0.45142180025577544, "num_tokens": 1651031.0, "step": 3740 }, { "entropy": 2.8970886945724486, "epoch": 1.1984, "grad_norm": 0.98046875, "learning_rate": 5.803687345576673e-05, "loss": 2.8164, "mean_token_accuracy": 0.45862447917461396, "num_tokens": 1653095.0, "step": 3745 }, { "entropy": 2.964561367034912, "epoch": 1.2, "grad_norm": 1.03125, "learning_rate": 5.7727791040927977e-05, "loss": 2.8807, "mean_token_accuracy": 0.45071675479412077, "num_tokens": 1655494.0, "step": 3750 }, { "entropy": 2.9308521509170533, "epoch": 1.2016, "grad_norm": 1.1796875, "learning_rate": 5.742466525345213e-05, "loss": 2.8898, "mean_token_accuracy": 0.4472826600074768, "num_tokens": 1657666.0, "step": 3755 }, { "entropy": 2.8843185901641846, "epoch": 1.2032, "grad_norm": 1.1171875, "learning_rate": 5.7127504403685775e-05, "loss": 2.8468, "mean_token_accuracy": 0.44261243641376496, "num_tokens": 1659842.0, "step": 3760 }, { "entropy": 2.9167650938034058, "epoch": 1.2048, "grad_norm": 1.0859375, "learning_rate": 5.6836316638443664e-05, "loss": 2.8481, "mean_token_accuracy": 0.4497327387332916, "num_tokens": 1661935.0, "step": 3765 }, { "entropy": 2.861691117286682, "epoch": 1.2064, "grad_norm": 1.1953125, "learning_rate": 5.655110994078553e-05, "loss": 2.8553, "mean_token_accuracy": 0.44176470935344697, "num_tokens": 1663978.0, "step": 3770 }, { "entropy": 2.9823601245880127, "epoch": 1.208, "grad_norm": 1.109375, "learning_rate": 5.6271892129797056e-05, "loss": 2.9346, "mean_token_accuracy": 0.4393131136894226, "num_tokens": 1666399.0, "step": 3775 }, { "entropy": 2.9265677690505982, "epoch": 1.2096, "grad_norm": 0.96875, "learning_rate": 5.599867086037556e-05, "loss": 2.8649, "mean_token_accuracy": 0.444303897023201, "num_tokens": 1668668.0, "step": 3780 }, { "entropy": 2.9490441560745237, "epoch": 1.2112, "grad_norm": 1.0859375, "learning_rate": 5.573145362302012e-05, "loss": 2.8446, "mean_token_accuracy": 0.4529551357030869, "num_tokens": 1670900.0, "step": 3785 }, { "entropy": 2.9436744451522827, "epoch": 1.2128, "grad_norm": 0.98046875, "learning_rate": 5.5470247743626404e-05, "loss": 2.8447, "mean_token_accuracy": 0.45734468996524813, "num_tokens": 1673141.0, "step": 3790 }, { "entropy": 2.9160199642181395, "epoch": 1.2144, "grad_norm": 1.1640625, "learning_rate": 5.5215060383285414e-05, "loss": 2.8781, "mean_token_accuracy": 0.43999702036380767, "num_tokens": 1675333.0, "step": 3795 }, { "entropy": 2.9387246131896974, "epoch": 1.216, "grad_norm": 1.0703125, "learning_rate": 5.496589853808759e-05, "loss": 2.8245, "mean_token_accuracy": 0.45650690197944643, "num_tokens": 1677544.0, "step": 3800 }, { "entropy": 2.8797014951705933, "epoch": 1.2176, "grad_norm": 1.0625, "learning_rate": 5.47227690389308e-05, "loss": 2.788, "mean_token_accuracy": 0.4717106282711029, "num_tokens": 1679520.0, "step": 3805 }, { "entropy": 2.9482503175735473, "epoch": 1.2192, "grad_norm": 1.03125, "learning_rate": 5.448567855133306e-05, "loss": 2.8886, "mean_token_accuracy": 0.44782242774963377, "num_tokens": 1681808.0, "step": 3810 }, { "entropy": 2.812504196166992, "epoch": 1.2208, "grad_norm": 1.515625, "learning_rate": 5.425463357524986e-05, "loss": 2.7602, "mean_token_accuracy": 0.4656171202659607, "num_tokens": 1683791.0, "step": 3815 }, { "entropy": 2.9265854358673096, "epoch": 1.2224, "grad_norm": 1.109375, "learning_rate": 5.402964044489591e-05, "loss": 2.8854, "mean_token_accuracy": 0.4432608515024185, "num_tokens": 1686106.0, "step": 3820 }, { "entropy": 2.8907620429992678, "epoch": 1.224, "grad_norm": 1.1171875, "learning_rate": 5.381070532857153e-05, "loss": 2.82, "mean_token_accuracy": 0.45292717814445493, "num_tokens": 1688240.0, "step": 3825 }, { "entropy": 2.923473334312439, "epoch": 1.2256, "grad_norm": 1.25, "learning_rate": 5.359783422849357e-05, "loss": 2.8609, "mean_token_accuracy": 0.4442437708377838, "num_tokens": 1690461.0, "step": 3830 }, { "entropy": 2.8852312564849854, "epoch": 1.2272, "grad_norm": 0.98046875, "learning_rate": 5.3391032980630736e-05, "loss": 2.8614, "mean_token_accuracy": 0.4556260943412781, "num_tokens": 1692669.0, "step": 3835 }, { "entropy": 2.8922724962234496, "epoch": 1.2288000000000001, "grad_norm": 1.0234375, "learning_rate": 5.31903072545437e-05, "loss": 2.8708, "mean_token_accuracy": 0.4496967136859894, "num_tokens": 1694803.0, "step": 3840 }, { "entropy": 2.8960639953613283, "epoch": 1.2304, "grad_norm": 0.98828125, "learning_rate": 5.29956625532297e-05, "loss": 2.8488, "mean_token_accuracy": 0.44592776000499723, "num_tokens": 1696943.0, "step": 3845 }, { "entropy": 2.9246564388275145, "epoch": 1.232, "grad_norm": 1.140625, "learning_rate": 5.280710421297146e-05, "loss": 2.875, "mean_token_accuracy": 0.45205327570438386, "num_tokens": 1699212.0, "step": 3850 }, { "entropy": 2.955587458610535, "epoch": 1.2336, "grad_norm": 1.0625, "learning_rate": 5.2624637403191165e-05, "loss": 2.8794, "mean_token_accuracy": 0.4469514340162277, "num_tokens": 1701442.0, "step": 3855 }, { "entropy": 2.9454181671142576, "epoch": 1.2352, "grad_norm": 0.99609375, "learning_rate": 5.2448267126308605e-05, "loss": 2.9011, "mean_token_accuracy": 0.4544678717851639, "num_tokens": 1703798.0, "step": 3860 }, { "entropy": 2.9179775714874268, "epoch": 1.2368000000000001, "grad_norm": 1.0234375, "learning_rate": 5.2277998217603954e-05, "loss": 2.909, "mean_token_accuracy": 0.44071272015571594, "num_tokens": 1706067.0, "step": 3865 }, { "entropy": 2.910015511512756, "epoch": 1.2384, "grad_norm": 1.140625, "learning_rate": 5.211383534508541e-05, "loss": 2.8787, "mean_token_accuracy": 0.44529192745685575, "num_tokens": 1708152.0, "step": 3870 }, { "entropy": 2.970757532119751, "epoch": 1.24, "grad_norm": 1.0390625, "learning_rate": 5.1955783009361044e-05, "loss": 2.8833, "mean_token_accuracy": 0.44263890087604524, "num_tokens": 1710423.0, "step": 3875 }, { "entropy": 2.978500247001648, "epoch": 1.2416, "grad_norm": 0.9453125, "learning_rate": 5.180384554351543e-05, "loss": 2.8901, "mean_token_accuracy": 0.4484690725803375, "num_tokens": 1712759.0, "step": 3880 }, { "entropy": 2.9426377773284913, "epoch": 1.2432, "grad_norm": 1.25, "learning_rate": 5.1658027112990976e-05, "loss": 2.8731, "mean_token_accuracy": 0.44753162264823915, "num_tokens": 1715023.0, "step": 3885 }, { "entropy": 2.958217406272888, "epoch": 1.2448, "grad_norm": 1.046875, "learning_rate": 5.151833171547365e-05, "loss": 2.852, "mean_token_accuracy": 0.4546335697174072, "num_tokens": 1717231.0, "step": 3890 }, { "entropy": 2.90898175239563, "epoch": 1.2464, "grad_norm": 0.984375, "learning_rate": 5.1384763180783274e-05, "loss": 2.8331, "mean_token_accuracy": 0.4459168463945389, "num_tokens": 1719347.0, "step": 3895 }, { "entropy": 2.9438149452209474, "epoch": 1.248, "grad_norm": 0.88671875, "learning_rate": 5.125732517076876e-05, "loss": 2.8717, "mean_token_accuracy": 0.4464673072099686, "num_tokens": 1721578.0, "step": 3900 }, { "entropy": 2.9742016553878785, "epoch": 1.2496, "grad_norm": 1.265625, "learning_rate": 5.113602117920747e-05, "loss": 2.9091, "mean_token_accuracy": 0.44329715967178346, "num_tokens": 1723952.0, "step": 3905 }, { "entropy": 2.8949837923049926, "epoch": 1.2511999999999999, "grad_norm": 1.140625, "learning_rate": 5.102085453170966e-05, "loss": 2.8205, "mean_token_accuracy": 0.4480966985225677, "num_tokens": 1726080.0, "step": 3910 }, { "entropy": 2.9691314697265625, "epoch": 1.2528000000000001, "grad_norm": 1.09375, "learning_rate": 5.091182838562709e-05, "loss": 2.9219, "mean_token_accuracy": 0.43697127401828767, "num_tokens": 1728485.0, "step": 3915 }, { "entropy": 2.931003284454346, "epoch": 1.2544, "grad_norm": 0.97265625, "learning_rate": 5.0808945729966716e-05, "loss": 2.8523, "mean_token_accuracy": 0.44951727986335754, "num_tokens": 1730714.0, "step": 3920 }, { "entropy": 2.9003567934036254, "epoch": 1.256, "grad_norm": 1.0625, "learning_rate": 5.071220938530844e-05, "loss": 2.8401, "mean_token_accuracy": 0.45560978055000306, "num_tokens": 1732831.0, "step": 3925 }, { "entropy": 2.915512299537659, "epoch": 1.2576, "grad_norm": 1.078125, "learning_rate": 5.0621622003728026e-05, "loss": 2.7965, "mean_token_accuracy": 0.4572917401790619, "num_tokens": 1734928.0, "step": 3930 }, { "entropy": 2.92191059589386, "epoch": 1.2591999999999999, "grad_norm": 1.2109375, "learning_rate": 5.053718606872433e-05, "loss": 2.8602, "mean_token_accuracy": 0.44150644838809966, "num_tokens": 1737073.0, "step": 3935 }, { "entropy": 2.959311008453369, "epoch": 1.2608, "grad_norm": 1.171875, "learning_rate": 5.0458903895151134e-05, "loss": 2.8441, "mean_token_accuracy": 0.44962053596973417, "num_tokens": 1739286.0, "step": 3940 }, { "entropy": 2.92020103931427, "epoch": 1.2624, "grad_norm": 1.1015625, "learning_rate": 5.038677762915381e-05, "loss": 2.8436, "mean_token_accuracy": 0.44944019615650177, "num_tokens": 1741524.0, "step": 3945 }, { "entropy": 2.966303491592407, "epoch": 1.264, "grad_norm": 0.99609375, "learning_rate": 5.032080924811033e-05, "loss": 2.9122, "mean_token_accuracy": 0.436164864897728, "num_tokens": 1743883.0, "step": 3950 }, { "entropy": 2.9542957544326782, "epoch": 1.2656, "grad_norm": 1.0703125, "learning_rate": 5.026100056057718e-05, "loss": 2.9017, "mean_token_accuracy": 0.4297113209962845, "num_tokens": 1746244.0, "step": 3955 }, { "entropy": 2.8869834184646606, "epoch": 1.2671999999999999, "grad_norm": 2.28125, "learning_rate": 5.0207353206239764e-05, "loss": 2.7956, "mean_token_accuracy": 0.4581418365240097, "num_tokens": 1748324.0, "step": 3960 }, { "entropy": 2.9147607564926146, "epoch": 1.2688, "grad_norm": 1.0, "learning_rate": 5.0159868655867436e-05, "loss": 2.8305, "mean_token_accuracy": 0.4662155359983444, "num_tokens": 1750484.0, "step": 3965 }, { "entropy": 2.891223835945129, "epoch": 1.2704, "grad_norm": 1.046875, "learning_rate": 5.011854821127305e-05, "loss": 2.8483, "mean_token_accuracy": 0.4486550480127335, "num_tokens": 1752647.0, "step": 3970 }, { "entropy": 2.903191065788269, "epoch": 1.272, "grad_norm": 1.1328125, "learning_rate": 5.008339300527755e-05, "loss": 2.8335, "mean_token_accuracy": 0.45128952860832217, "num_tokens": 1754863.0, "step": 3975 }, { "entropy": 2.8733340978622435, "epoch": 1.2736, "grad_norm": 1.0390625, "learning_rate": 5.005440400167859e-05, "loss": 2.8089, "mean_token_accuracy": 0.4564253330230713, "num_tokens": 1757004.0, "step": 3980 }, { "entropy": 2.9120566844940186, "epoch": 1.2752, "grad_norm": 1.046875, "learning_rate": 5.003158199522442e-05, "loss": 2.8531, "mean_token_accuracy": 0.4559608221054077, "num_tokens": 1759220.0, "step": 3985 }, { "entropy": 2.922964310646057, "epoch": 1.2768, "grad_norm": 1.1171875, "learning_rate": 5.0014927611591806e-05, "loss": 2.8901, "mean_token_accuracy": 0.4424146503210068, "num_tokens": 1761510.0, "step": 3990 }, { "entropy": 2.926527261734009, "epoch": 1.2784, "grad_norm": 1.0546875, "learning_rate": 5.000444130736916e-05, "loss": 2.8799, "mean_token_accuracy": 0.451074543595314, "num_tokens": 1763819.0, "step": 3995 }, { "entropy": 2.8954465627670287, "epoch": 1.28, "grad_norm": 0.99609375, "learning_rate": 5.0000123370043736e-05, "loss": 2.8555, "mean_token_accuracy": 0.45639456510543824, "num_tokens": 1765972.0, "step": 4000 } ], "logging_steps": 5, "max_steps": 4000, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2466907066368000.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }