{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.48, "eval_steps": 500, "global_step": 1500, "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 } ], "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": false }, "attributes": {} } }, "total_flos": 933743407104000.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }