BananaMind-2.1-MiniCoder / training_metrics.jsonl
Banaxi-Tech's picture
Save 15% checkpoint at 4,500,488,192 tokens
bb2934d verified
Raw
History Blame Contribute Delete
435 kB
{"step": 1, "total_steps": 57221, "tokens": 524288, "source": "stack_v3", "source_tokens": {"stack_v3": 524288, "fineweb_edu": 0}, "loss": 7.477823257446289, "ce_loss": 7.4691489934921265, "perplexity": 1753.1141390404482, "z_loss": 86.7425651550293, "grad_norm": 3.236438274383545, "muon_lr": 1.7467248908296942e-05, "adamw_lr": 1.7467248908296944e-06, "ngram_lr": 8.733624454148472e-07, "weight_decay": 0.1, "tokens_per_second": 4567.832061431627, "data_wait_seconds": 19.332803905010223, "eta_seconds": 6567552.245473253}
{"step": 10, "total_steps": 57221, "tokens": 5242880, "source": "stack_v3", "source_tokens": {"stack_v3": 4194304, "fineweb_edu": 1048576}, "loss": 7.311760187149048, "ce_loss": 7.303089737892151, "perplexity": 1484.880739472366, "z_loss": 86.70331001281738, "grad_norm": 3.5776402950286865, "muon_lr": 0.00017467248908296945, "adamw_lr": 1.7467248908296942e-05, "ngram_lr": 8.733624454148471e-06, "weight_decay": 0.1, "tokens_per_second": 450070.2125749802, "data_wait_seconds": 2.2172927856445312e-05, "eta_seconds": 66644.61739956401}
{"step": 20, "total_steps": 57221, "tokens": 10485760, "source": "stack_v3", "source_tokens": {"stack_v3": 7864320, "fineweb_edu": 2621440}, "loss": 7.265507340431213, "ce_loss": 7.256851553916931, "perplexity": 1417.7856809955563, "z_loss": 86.55733299255371, "grad_norm": 3.5466527938842773, "muon_lr": 0.0003493449781659389, "adamw_lr": 3.4934497816593884e-05, "ngram_lr": 1.7467248908296942e-05, "weight_decay": 0.1, "tokens_per_second": 1127545.278643539, "data_wait_seconds": 2.1696090698242188e-05, "eta_seconds": 26597.170692850603}
{"step": 30, "total_steps": 57221, "tokens": 15728640, "source": "stack_v3", "source_tokens": {"stack_v3": 12058624, "fineweb_edu": 3670016}, "loss": 7.502801418304443, "ce_loss": 7.4941771030426025, "perplexity": 1797.544962234031, "z_loss": 86.24256134033203, "grad_norm": 2.5389275550842285, "muon_lr": 0.0005240174672489083, "adamw_lr": 5.2401746724890826e-05, "ngram_lr": 2.6200873362445413e-05, "weight_decay": 0.1, "tokens_per_second": 1128254.0694329138, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 26575.814944829563}
{"step": 40, "total_steps": 57221, "tokens": 20971520, "source": "stack_v3", "source_tokens": {"stack_v3": 15728640, "fineweb_edu": 5242880}, "loss": 7.030381202697754, "ce_loss": 7.021787405014038, "perplexity": 1120.7881307810303, "z_loss": 85.93887710571289, "grad_norm": 4.1022868156433105, "muon_lr": 0.0006986899563318778, "adamw_lr": 6.986899563318777e-05, "ngram_lr": 3.4934497816593884e-05, "weight_decay": 0.1, "tokens_per_second": 481176.3895548848, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 62303.61491288524}
{"step": 50, "total_steps": 57221, "tokens": 26214400, "source": "stack_v3", "source_tokens": {"stack_v3": 19922944, "fineweb_edu": 6291456}, "loss": 7.8793723583221436, "ce_loss": 7.870861291885376, "perplexity": 2619.82104735295, "z_loss": 85.10893440246582, "grad_norm": 4.983921051025391, "muon_lr": 0.0008733624454148472, "adamw_lr": 8.733624454148472e-05, "ngram_lr": 4.366812227074236e-05, "weight_decay": 0.1, "tokens_per_second": 537835.9967189172, "data_wait_seconds": 2.759695053100586e-05, "eta_seconds": 55730.344905985985}
{"step": 60, "total_steps": 57221, "tokens": 31457280, "source": "stack_v3", "source_tokens": {"stack_v3": 23592960, "fineweb_edu": 7864320}, "loss": 7.152419924736023, "ce_loss": 7.1440346240997314, "perplexity": 1266.5280591851786, "z_loss": 83.85223197937012, "grad_norm": 6.806609153747559, "muon_lr": 0.0010480349344978166, "adamw_lr": 0.00010480349344978165, "ngram_lr": 5.2401746724890826e-05, "weight_decay": 0.1, "tokens_per_second": 1171019.587586067, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 25591.83726531593}
{"step": 70, "total_steps": 57221, "tokens": 36700160, "source": "stack_v3", "source_tokens": {"stack_v3": 27787264, "fineweb_edu": 8912896}, "loss": 6.888280749320984, "ce_loss": 6.880123615264893, "perplexity": 972.7465986885198, "z_loss": 81.57044792175293, "grad_norm": 8.526148796081543, "muon_lr": 0.001222707423580786, "adamw_lr": 0.0001222707423580786, "ngram_lr": 6.11353711790393e-05, "weight_decay": 0.1, "tokens_per_second": 1144395.327498384, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 26182.64783158363}
{"step": 80, "total_steps": 57221, "tokens": 41943040, "source": "stack_v3", "source_tokens": {"stack_v3": 31457280, "fineweb_edu": 10485760}, "loss": 6.238178133964539, "ce_loss": 6.230307936668396, "perplexity": 507.91186410416725, "z_loss": 78.7029037475586, "grad_norm": 8.414324760437012, "muon_lr": 0.0013973799126637556, "adamw_lr": 0.00013973799126637554, "ngram_lr": 6.986899563318777e-05, "weight_decay": 0.1, "tokens_per_second": 1126133.1335191084, "data_wait_seconds": 1.5556812286376953e-05, "eta_seconds": 26602.58904414135}
{"step": 90, "total_steps": 57221, "tokens": 47185920, "source": "stack_v3", "source_tokens": {"stack_v3": 35651584, "fineweb_edu": 11534336}, "loss": 5.654323220252991, "ce_loss": 5.646766781806946, "perplexity": 283.37377383314856, "z_loss": 75.56427383422852, "grad_norm": 6.446281909942627, "muon_lr": 0.0015720524017467249, "adamw_lr": 0.0001572052401746725, "ngram_lr": 7.860262008733625e-05, "weight_decay": 0.1, "tokens_per_second": 1169351.7454488024, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 25614.888075019702}
{"step": 100, "total_steps": 57221, "tokens": 52428800, "source": "stack_v3", "source_tokens": {"stack_v3": 39321600, "fineweb_edu": 13107200}, "loss": 5.2927528619766235, "ce_loss": 5.285556197166443, "perplexity": 197.46398181771718, "z_loss": 71.96705627441406, "grad_norm": 4.47438907623291, "muon_lr": 0.0017467248908296944, "adamw_lr": 0.00017467248908296945, "ngram_lr": 8.733624454148472e-05, "weight_decay": 0.1, "tokens_per_second": 1175082.2231187688, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 25485.511235559825}
{"step": 110, "total_steps": 57221, "tokens": 57671680, "source": "stack_v3", "source_tokens": {"stack_v3": 43515904, "fineweb_edu": 14155776}, "loss": 4.0975231528282166, "ce_loss": 4.090309679508209, "perplexity": 59.7583947892939, "z_loss": 72.13484764099121, "grad_norm": 2.999368667602539, "muon_lr": 0.0019213973799126639, "adamw_lr": 0.00019213973799126638, "ngram_lr": 9.606986899563319e-05, "weight_decay": 0.1, "tokens_per_second": 1168761.2582711307, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 25618.857664987678}
{"step": 120, "total_steps": 57221, "tokens": 62914560, "source": "stack_v3", "source_tokens": {"stack_v3": 47185920, "fineweb_edu": 15728640}, "loss": 4.131748557090759, "ce_loss": 4.124546647071838, "perplexity": 61.83976765473531, "z_loss": 72.02021789550781, "grad_norm": 2.474316120147705, "muon_lr": 0.002096069868995633, "adamw_lr": 0.0002096069868995633, "ngram_lr": 0.00010480349344978165, "weight_decay": 0.1, "tokens_per_second": 1132060.9265514386, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 26444.765239973785}
{"step": 130, "total_steps": 57221, "tokens": 68157440, "source": "stack_v3", "source_tokens": {"stack_v3": 51380224, "fineweb_edu": 16777216}, "loss": 3.899881064891815, "ce_loss": 3.892593502998352, "perplexity": 49.03790169049644, "z_loss": 72.87546920776367, "grad_norm": 1.4456034898757935, "muon_lr": 0.002270742358078603, "adamw_lr": 0.00022707423580786026, "ngram_lr": 0.00011353711790393013, "weight_decay": 0.1, "tokens_per_second": 1173154.7138645095, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 25513.977147482103}
{"step": 140, "total_steps": 57221, "tokens": 73400320, "source": "stack_v3", "source_tokens": {"stack_v3": 55050240, "fineweb_edu": 18350080}, "loss": 3.2269490361213684, "ce_loss": 3.2192242443561554, "perplexity": 25.00871200482686, "z_loss": 77.24747848510742, "grad_norm": 1.3326994180679321, "muon_lr": 0.002445414847161572, "adamw_lr": 0.0002445414847161572, "ngram_lr": 0.0001222707423580786, "weight_decay": 0.1, "tokens_per_second": 1186641.619323998, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 25219.57699161815}
{"step": 150, "total_steps": 57221, "tokens": 78643200, "source": "stack_v3", "source_tokens": {"stack_v3": 59244544, "fineweb_edu": 19398656}, "loss": 3.3159977197647095, "ce_loss": 3.3083859980106354, "perplexity": 27.34096147488936, "z_loss": 76.11794662475586, "grad_norm": 1.3569964170455933, "muon_lr": 0.0026200873362445414, "adamw_lr": 0.00026200873362445414, "ngram_lr": 0.00013100436681222707, "weight_decay": 0.1, "tokens_per_second": 1169444.4030801957, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 25585.959213785827}
{"step": 160, "total_steps": 57221, "tokens": 83886080, "source": "stack_v3", "source_tokens": {"stack_v3": 62914560, "fineweb_edu": 20971520}, "loss": 3.3972354233264923, "ce_loss": 3.3895267248153687, "perplexity": 29.65191543323618, "z_loss": 77.08673095703125, "grad_norm": 2.3261313438415527, "muon_lr": 0.002794759825327511, "adamw_lr": 0.00027947598253275107, "ngram_lr": 0.00013973799126637554, "weight_decay": 0.1, "tokens_per_second": 564989.6002295082, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 52949.84882526612}
{"step": 170, "total_steps": 57221, "tokens": 89128960, "source": "stack_v3", "source_tokens": {"stack_v3": 67108864, "fineweb_edu": 22020096}, "loss": 2.736926347017288, "ce_loss": 2.7285077571868896, "perplexity": 15.310023692508153, "z_loss": 84.18612098693848, "grad_norm": 0.8374522924423218, "muon_lr": 0.0029694323144104804, "adamw_lr": 0.00029694323144104805, "ngram_lr": 0.00014847161572052403, "weight_decay": 0.1, "tokens_per_second": 497318.8050426191, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 60144.25904815062}
{"step": 180, "total_steps": 57221, "tokens": 94371840, "source": "stack_v3", "source_tokens": {"stack_v3": 70778880, "fineweb_edu": 23592960}, "loss": 3.0876591205596924, "ce_loss": 3.079368159174919, "perplexity": 21.744658891585498, "z_loss": 82.9090461730957, "grad_norm": 0.9704871773719788, "muon_lr": 0.0031441048034934497, "adamw_lr": 0.000314410480349345, "ngram_lr": 0.0001572052401746725, "weight_decay": 0.1, "tokens_per_second": 1179571.5682337387, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 25352.957773287082}
{"step": 190, "total_steps": 57221, "tokens": 99614720, "source": "stack_v3", "source_tokens": {"stack_v3": 74973184, "fineweb_edu": 24641536}, "loss": 2.955415725708008, "ce_loss": 2.9469285011291504, "perplexity": 19.04735984451778, "z_loss": 84.87235069274902, "grad_norm": 0.7044450640678406, "muon_lr": 0.0033187772925764194, "adamw_lr": 0.0003318777292576419, "ngram_lr": 0.00016593886462882096, "weight_decay": 0.1, "tokens_per_second": 1138356.3889315242, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 26266.277916764608}
{"step": 200, "total_steps": 57221, "tokens": 104857600, "source": "stack_v3", "source_tokens": {"stack_v3": 78643200, "fineweb_edu": 26214400}, "loss": 3.160772427916527, "ce_loss": 3.1520421355962753, "perplexity": 23.383768681643065, "z_loss": 87.30314445495605, "grad_norm": 0.6705701947212219, "muon_lr": 0.0034934497816593887, "adamw_lr": 0.0003493449781659389, "ngram_lr": 0.00017467248908296945, "weight_decay": 0.1, "tokens_per_second": 1119900.2517849335, "data_wait_seconds": 2.1576881408691406e-05, "eta_seconds": 26694.46886216174}
{"step": 210, "total_steps": 57221, "tokens": 110100480, "source": "stack_v3", "source_tokens": {"stack_v3": 82837504, "fineweb_edu": 27262976}, "loss": 2.511118784546852, "ce_loss": 2.5018782913684845, "perplexity": 12.205397737205663, "z_loss": 92.4043960571289, "grad_norm": 0.6518954038619995, "muon_lr": 0.0036681222707423584, "adamw_lr": 0.00036681222707423577, "ngram_lr": 0.00018340611353711788, "weight_decay": 0.1, "tokens_per_second": 1165444.0615831302, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 25646.79035679996}
{"step": 220, "total_steps": 57221, "tokens": 115343360, "source": "stack_v3", "source_tokens": {"stack_v3": 86507520, "fineweb_edu": 28835840}, "loss": 2.1069177389144897, "ce_loss": 2.0975685715675354, "perplexity": 8.146338573868261, "z_loss": 93.49156188964844, "grad_norm": 0.8171663284301758, "muon_lr": 0.0038427947598253277, "adamw_lr": 0.00038427947598253275, "ngram_lr": 0.00019213973799126638, "weight_decay": 0.1, "tokens_per_second": 1090748.7703051227, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 27398.295055276714}
{"step": 230, "total_steps": 57221, "tokens": 120586240, "source": "stack_v3", "source_tokens": {"stack_v3": 90701824, "fineweb_edu": 29884416}, "loss": 2.2503341138362885, "ce_loss": 2.2409065663814545, "perplexity": 9.401850826978416, "z_loss": 94.27532768249512, "grad_norm": 0.8574920892715454, "muon_lr": 0.004017467248908297, "adamw_lr": 0.00040174672489082973, "ngram_lr": 0.00020087336244541487, "weight_decay": 0.1, "tokens_per_second": 1158663.6063405364, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 25787.824521708768}
{"step": 240, "total_steps": 57221, "tokens": 125829120, "source": "stack_v3", "source_tokens": {"stack_v3": 94371840, "fineweb_edu": 31457280}, "loss": 2.058788627386093, "ce_loss": 2.0487880408763885, "perplexity": 7.7584924303066884, "z_loss": 100.0061149597168, "grad_norm": 0.6514208316802979, "muon_lr": 0.004192139737991266, "adamw_lr": 0.0004192139737991266, "ngram_lr": 0.0002096069868995633, "weight_decay": 0.1, "tokens_per_second": 1116492.0766467005, "data_wait_seconds": 2.390146255493164e-05, "eta_seconds": 26757.172312162584}
{"step": 250, "total_steps": 57221, "tokens": 131072000, "source": "stack_v3", "source_tokens": {"stack_v3": 98566144, "fineweb_edu": 32505856}, "loss": 2.600365087389946, "ce_loss": 2.5905019342899323, "perplexity": 13.336463952059566, "z_loss": 98.63163566589355, "grad_norm": 0.5599244832992554, "muon_lr": 0.004366812227074236, "adamw_lr": 0.0004366812227074236, "ngram_lr": 0.0002183406113537118, "weight_decay": 0.1, "tokens_per_second": 1169598.9065732784, "data_wait_seconds": 1.7404556274414062e-05, "eta_seconds": 25537.75301270653}
{"step": 260, "total_steps": 57221, "tokens": 136314880, "source": "stack_v3", "source_tokens": {"stack_v3": 102236160, "fineweb_edu": 34078720}, "loss": 2.4170388281345367, "ce_loss": 2.407380387187004, "perplexity": 11.10483264714103, "z_loss": 96.58390045166016, "grad_norm": 0.5581417083740234, "muon_lr": 0.004541484716157206, "adamw_lr": 0.0004541484716157205, "ngram_lr": 0.00022707423580786026, "weight_decay": 0.1, "tokens_per_second": 1174941.2712174437, "data_wait_seconds": 1.7464160919189453e-05, "eta_seconds": 25417.17262945068}
{"step": 270, "total_steps": 57221, "tokens": 141557760, "source": "stack_v3", "source_tokens": {"stack_v3": 106430464, "fineweb_edu": 35127296}, "loss": 3.077435612678528, "ce_loss": 3.0687750577926636, "perplexity": 21.515531244226278, "z_loss": 86.60528945922852, "grad_norm": 0.6072813868522644, "muon_lr": 0.004716157205240175, "adamw_lr": 0.0004716157205240175, "ngram_lr": 0.00023580786026200875, "weight_decay": 0.1, "tokens_per_second": 1089321.2453896292, "data_wait_seconds": 2.3543834686279297e-05, "eta_seconds": 27410.134858170524}
{"step": 280, "total_steps": 57221, "tokens": 146800640, "source": "stack_v3", "source_tokens": {"stack_v3": 110100480, "fineweb_edu": 36700160}, "loss": 3.734122633934021, "ce_loss": 3.7255266308784485, "perplexity": 41.49307852838871, "z_loss": 85.95946884155273, "grad_norm": 0.7921918630599976, "muon_lr": 0.004890829694323144, "adamw_lr": 0.0004890829694323144, "ngram_lr": 0.0002445414847161572, "weight_decay": 0.1, "tokens_per_second": 1089859.7210999667, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 27391.781512826212}
{"step": 290, "total_steps": 57221, "tokens": 152043520, "source": "stack_v3", "source_tokens": {"stack_v3": 114294784, "fineweb_edu": 37748736}, "loss": 3.3451258838176727, "ce_loss": 3.3362261056900024, "perplexity": 28.112831403476783, "z_loss": 88.99772644042969, "grad_norm": 1.0344287157058716, "muon_lr": 0.005065502183406114, "adamw_lr": 0.0005065502183406114, "ngram_lr": 0.0002532751091703057, "weight_decay": 0.1, "tokens_per_second": 1120162.8378380993, "data_wait_seconds": 2.962350845336914e-05, "eta_seconds": 26646.087043564305}
{"step": 300, "total_steps": 57221, "tokens": 157286400, "source": "stack_v3", "source_tokens": {"stack_v3": 117964800, "fineweb_edu": 39321600}, "loss": 4.055304944515228, "ce_loss": 4.0468446016311646, "perplexity": 57.21663064281139, "z_loss": 84.6041374206543, "grad_norm": 0.625637948513031, "muon_lr": 0.005240174672489083, "adamw_lr": 0.0005240174672489083, "ngram_lr": 0.00026200873362445414, "weight_decay": 0.1, "tokens_per_second": 1175718.9055054928, "data_wait_seconds": 2.282857894897461e-05, "eta_seconds": 25382.524224333465}
{"step": 310, "total_steps": 57221, "tokens": 162529280, "source": "stack_v3", "source_tokens": {"stack_v3": 122159104, "fineweb_edu": 40370176}, "loss": 2.8337939381599426, "ce_loss": 2.824828326702118, "perplexity": 16.85805066050418, "z_loss": 89.65686988830566, "grad_norm": 0.7494619488716125, "muon_lr": 0.005414847161572052, "adamw_lr": 0.0005414847161572052, "ngram_lr": 0.0002707423580786026, "weight_decay": 0.1, "tokens_per_second": 444770.5318897706, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 67085.08900808822}
{"step": 320, "total_steps": 57221, "tokens": 167772160, "source": "stack_v3", "source_tokens": {"stack_v3": 125829120, "fineweb_edu": 41943040}, "loss": 2.574665606021881, "ce_loss": 2.565473258495331, "perplexity": 13.006812497820828, "z_loss": 91.92424774169922, "grad_norm": 0.6182001829147339, "muon_lr": 0.005589519650655022, "adamw_lr": 0.0005589519650655021, "ngram_lr": 0.00027947598253275107, "weight_decay": 0.1, "tokens_per_second": 1128129.220142111, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 26443.981156912167}
{"step": 330, "total_steps": 57221, "tokens": 173015040, "source": "stack_v3", "source_tokens": {"stack_v3": 130023424, "fineweb_edu": 42991616}, "loss": 2.8070927560329437, "ce_loss": 2.797764629125595, "perplexity": 16.407927942055526, "z_loss": 93.28130722045898, "grad_norm": 0.45890071988105774, "muon_lr": 0.005764192139737992, "adamw_lr": 0.0005764192139737992, "ngram_lr": 0.0002882096069868996, "weight_decay": 0.1, "tokens_per_second": 1181553.1233626206, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 25243.879746273626}
{"step": 340, "total_steps": 57221, "tokens": 178257920, "source": "stack_v3", "source_tokens": {"stack_v3": 133693440, "fineweb_edu": 44564480}, "loss": 3.01384699344635, "ce_loss": 3.0043631196022034, "perplexity": 20.173363983713802, "z_loss": 94.83857727050781, "grad_norm": 0.7018427848815918, "muon_lr": 0.005938864628820961, "adamw_lr": 0.0005938864628820961, "ngram_lr": 0.00029694323144104805, "weight_decay": 0.1, "tokens_per_second": 1090689.9096487788, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 27342.090374342162}
{"step": 350, "total_steps": 57221, "tokens": 183500800, "source": "stack_v3", "source_tokens": {"stack_v3": 137887744, "fineweb_edu": 45613056}, "loss": 2.631975084543228, "ce_loss": 2.622410476207733, "perplexity": 13.768873157052015, "z_loss": 95.64596748352051, "grad_norm": 0.6670660376548767, "muon_lr": 0.00611353711790393, "adamw_lr": 0.000611353711790393, "ngram_lr": 0.0003056768558951965, "weight_decay": 0.1, "tokens_per_second": 1150840.4346268578, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 25908.45637924387}
{"step": 360, "total_steps": 57221, "tokens": 188743680, "source": "stack_v3", "source_tokens": {"stack_v3": 141557760, "fineweb_edu": 47185920}, "loss": 3.007331132888794, "ce_loss": 2.997472822666168, "perplexity": 20.034841294940023, "z_loss": 98.58303260803223, "grad_norm": 0.7906071543693542, "muon_lr": 0.0062882096069868994, "adamw_lr": 0.00062882096069869, "ngram_lr": 0.000314410480349345, "weight_decay": 0.1, "tokens_per_second": 1133893.723339148, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 26291.04977511498}
{"step": 370, "total_steps": 57221, "tokens": 193986560, "source": "stack_v3", "source_tokens": {"stack_v3": 145752064, "fineweb_edu": 48234496}, "loss": 2.876849353313446, "ce_loss": 2.867273688316345, "perplexity": 17.588999677526047, "z_loss": 95.75677108764648, "grad_norm": 0.8373186588287354, "muon_lr": 0.00646288209606987, "adamw_lr": 0.0006462882096069869, "ngram_lr": 0.00032314410480349345, "weight_decay": 0.1, "tokens_per_second": 1155211.4169781606, "data_wait_seconds": 1.7583370208740234e-05, "eta_seconds": 25801.349434346426}
{"step": 380, "total_steps": 57221, "tokens": 199229440, "source": "stack_v3", "source_tokens": {"stack_v3": 149422080, "fineweb_edu": 49807360}, "loss": 2.648943841457367, "ce_loss": 2.6380085349082947, "perplexity": 13.985324571203739, "z_loss": 109.35313034057617, "grad_norm": 0.5392980575561523, "muon_lr": 0.006637554585152839, "adamw_lr": 0.0006637554585152838, "ngram_lr": 0.0003318777292576419, "weight_decay": 0.1, "tokens_per_second": 1139297.1370278636, "data_wait_seconds": 2.2649765014648438e-05, "eta_seconds": 26157.15390784061}
{"step": 390, "total_steps": 57221, "tokens": 204472320, "source": "stack_v3", "source_tokens": {"stack_v3": 153616384, "fineweb_edu": 50855936}, "loss": 2.2994812428951263, "ce_loss": 2.2883530855178833, "perplexity": 9.858687887978826, "z_loss": 111.28175163269043, "grad_norm": 0.3070160746574402, "muon_lr": 0.006812227074235807, "adamw_lr": 0.0006812227074235809, "ngram_lr": 0.00034061135371179043, "weight_decay": 0.1, "tokens_per_second": 1148418.82342054, "data_wait_seconds": 1.7702579498291016e-05, "eta_seconds": 25944.826985032065}
{"step": 400, "total_steps": 57221, "tokens": 209715200, "source": "stack_v3", "source_tokens": {"stack_v3": 157286400, "fineweb_edu": 52428800}, "loss": 2.803308844566345, "ce_loss": 2.7931976318359375, "perplexity": 16.333163833305374, "z_loss": 101.11209678649902, "grad_norm": 0.6259375214576721, "muon_lr": 0.0069868995633187774, "adamw_lr": 0.0006986899563318778, "ngram_lr": 0.0003493449781659389, "weight_decay": 0.1, "tokens_per_second": 1162147.5453853717, "data_wait_seconds": 2.3603439331054688e-05, "eta_seconds": 25633.823276821062}
{"step": 410, "total_steps": 57221, "tokens": 214958080, "source": "stack_v3", "source_tokens": {"stack_v3": 161480704, "fineweb_edu": 53477376}, "loss": 2.4341437816619873, "ce_loss": 2.4236778020858765, "perplexity": 11.287295510083888, "z_loss": 104.65981483459473, "grad_norm": 0.7161032557487488, "muon_lr": 0.007161572052401746, "adamw_lr": 0.0007161572052401747, "ngram_lr": 0.00035807860262008736, "weight_decay": 0.1, "tokens_per_second": 765467.7432039175, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 38910.90406413819}
{"step": 420, "total_steps": 57221, "tokens": 220200960, "source": "stack_v3", "source_tokens": {"stack_v3": 165150720, "fineweb_edu": 55050240}, "loss": 2.093996673822403, "ce_loss": 2.0831812620162964, "perplexity": 8.029973774525564, "z_loss": 108.15422058105469, "grad_norm": 0.8360431790351868, "muon_lr": 0.007336244541484717, "adamw_lr": 0.0007336244541484715, "ngram_lr": 0.00036681222707423577, "weight_decay": 0.1, "tokens_per_second": 1181115.9314268208, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 25213.273521783063}
{"step": 430, "total_steps": 57221, "tokens": 225443840, "source": "stack_v3", "source_tokens": {"stack_v3": 169345024, "fineweb_edu": 56098816}, "loss": 2.1151992082595825, "ce_loss": 2.103895843029022, "perplexity": 8.198046080652, "z_loss": 113.03414916992188, "grad_norm": 0.4469183683395386, "muon_lr": 0.007510917030567685, "adamw_lr": 0.0007510917030567686, "ngram_lr": 0.0003755458515283843, "weight_decay": 0.1, "tokens_per_second": 1146650.3072416745, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 25966.553160940763}
{"step": 440, "total_steps": 57221, "tokens": 230686720, "source": "stack_v3", "source_tokens": {"stack_v3": 173015040, "fineweb_edu": 57671680}, "loss": 1.8253012299537659, "ce_loss": 1.8138299882411957, "perplexity": 6.133895247810853, "z_loss": 114.71238327026367, "grad_norm": 0.7037718892097473, "muon_lr": 0.0076855895196506555, "adamw_lr": 0.0007685589519650655, "ngram_lr": 0.00038427947598253275, "weight_decay": 0.1, "tokens_per_second": 1155588.28257014, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 25761.17612908827}
{"step": 450, "total_steps": 57221, "tokens": 235929600, "source": "stack_v3", "source_tokens": {"stack_v3": 177209344, "fineweb_edu": 58720256}, "loss": 1.8191999644041061, "ce_loss": 1.8076701164245605, "perplexity": 6.096227373093412, "z_loss": 115.29867553710938, "grad_norm": 0.9965146780014038, "muon_lr": 0.007860262008733625, "adamw_lr": 0.0007860262008733624, "ngram_lr": 0.0003930131004366812, "weight_decay": 0.1, "tokens_per_second": 1073723.788795735, "data_wait_seconds": 1.6510486602783203e-05, "eta_seconds": 27720.41628451087}
{"step": 460, "total_steps": 57221, "tokens": 241172480, "source": "stack_v3", "source_tokens": {"stack_v3": 180879360, "fineweb_edu": 60293120}, "loss": 1.9125119745731354, "ce_loss": 1.9008100777864456, "perplexity": 6.691312731170914, "z_loss": 117.0188102722168, "grad_norm": 0.8753389120101929, "muon_lr": 0.008034934497816594, "adamw_lr": 0.0008034934497816595, "ngram_lr": 0.00040174672489082973, "weight_decay": 0.1, "tokens_per_second": 1140423.0502379304, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 26094.551064880103}
{"step": 470, "total_steps": 57221, "tokens": 246415360, "source": "stack_v3", "source_tokens": {"stack_v3": 185073664, "fineweb_edu": 61341696}, "loss": 1.4722235649824142, "ce_loss": 1.4596999287605286, "perplexity": 4.304667627573249, "z_loss": 125.23627471923828, "grad_norm": 0.4803929328918457, "muon_lr": 0.008209606986899563, "adamw_lr": 0.0008209606986899564, "ngram_lr": 0.0004104803493449782, "weight_decay": 0.1, "tokens_per_second": 481923.93562055216, "data_wait_seconds": 1.6391277313232422e-05, "eta_seconds": 61739.17176719522}
{"step": 480, "total_steps": 57221, "tokens": 251658240, "source": "stack_v3", "source_tokens": {"stack_v3": 188743680, "fineweb_edu": 62914560}, "loss": 2.066705659031868, "ce_loss": 2.055333688855171, "perplexity": 7.809443362115351, "z_loss": 113.71966743469238, "grad_norm": 0.9224481582641602, "muon_lr": 0.008384279475982533, "adamw_lr": 0.0008384279475982532, "ngram_lr": 0.0004192139737991266, "weight_decay": 0.1, "tokens_per_second": 1156230.8196937835, "data_wait_seconds": 1.7583370208740234e-05, "eta_seconds": 25728.722373858327}
{"step": 490, "total_steps": 57221, "tokens": 256901120, "source": "stack_v3", "source_tokens": {"stack_v3": 192937984, "fineweb_edu": 63963136}, "loss": 2.5063512325286865, "ce_loss": 2.4953786730766296, "perplexity": 12.126324561912401, "z_loss": 109.72551155090332, "grad_norm": 0.4465581774711609, "muon_lr": 0.008558951965065504, "adamw_lr": 0.0008558951965065503, "ngram_lr": 0.0004279475982532751, "weight_decay": 0.1, "tokens_per_second": 1160091.060937161, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 25638.589832743397}
{"step": 500, "total_steps": 57221, "tokens": 262144000, "source": "stack_v3", "source_tokens": {"stack_v3": 196608000, "fineweb_edu": 65536000}, "loss": 2.3740538954734802, "ce_loss": 2.362781375646591, "perplexity": 10.62044987095248, "z_loss": 112.72533988952637, "grad_norm": 0.48989182710647583, "muon_lr": 0.008733624454148471, "adamw_lr": 0.0008733624454148472, "ngram_lr": 0.0004366812227074236, "weight_decay": 0.1, "tokens_per_second": 1106799.9051514221, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 26868.321782094426}
{"step": 510, "total_steps": 57221, "tokens": 267386880, "source": "stack_v3", "source_tokens": {"stack_v3": 200802304, "fineweb_edu": 66584576}, "loss": 2.578342616558075, "ce_loss": 2.5678288638591766, "perplexity": 13.03748753005814, "z_loss": 105.13729858398438, "grad_norm": 0.7434462308883667, "muon_lr": 0.008908296943231442, "adamw_lr": 0.0008908296943231441, "ngram_lr": 0.00044541484716157205, "weight_decay": 0.1, "tokens_per_second": 1159681.1630651099, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 25638.610048140166}
{"step": 520, "total_steps": 57221, "tokens": 272629760, "source": "stack_v3", "source_tokens": {"stack_v3": 204472320, "fineweb_edu": 68157440}, "loss": 2.3549944162368774, "ce_loss": 2.344226837158203, "perplexity": 10.425209226359073, "z_loss": 107.67593574523926, "grad_norm": 0.5201312899589539, "muon_lr": 0.009082969432314411, "adamw_lr": 0.000908296943231441, "ngram_lr": 0.0004541484716157205, "weight_decay": 0.1, "tokens_per_second": 1126041.388087622, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 26399.891295724552}
{"step": 530, "total_steps": 57221, "tokens": 277872640, "source": "stack_v3", "source_tokens": {"stack_v3": 208666624, "fineweb_edu": 69206016}, "loss": 2.4058909118175507, "ce_loss": 2.3950898349285126, "perplexity": 10.969183430628696, "z_loss": 108.01037406921387, "grad_norm": 0.9829322695732117, "muon_lr": 0.009257641921397379, "adamw_lr": 0.0009257641921397381, "ngram_lr": 0.00046288209606986904, "weight_decay": 0.1, "tokens_per_second": 1118829.9612433056, "data_wait_seconds": 2.3484230041503906e-05, "eta_seconds": 26565.365953349276}
{"step": 540, "total_steps": 57221, "tokens": 283115520, "source": "stack_v3", "source_tokens": {"stack_v3": 212336640, "fineweb_edu": 70778880}, "loss": 2.276112735271454, "ce_loss": 2.264992266893387, "perplexity": 9.631050122884915, "z_loss": 111.20458602905273, "grad_norm": 1.2314324378967285, "muon_lr": 0.00943231441048035, "adamw_lr": 0.000943231441048035, "ngram_lr": 0.0004716157205240175, "weight_decay": 0.1, "tokens_per_second": 1144346.2558193086, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 25968.437724929536}
{"step": 550, "total_steps": 57221, "tokens": 288358400, "source": "stack_v3", "source_tokens": {"stack_v3": 216530944, "fineweb_edu": 71827456}, "loss": 2.5301115810871124, "ce_loss": 2.5194147527217865, "perplexity": 12.42132498927952, "z_loss": 106.96755981445312, "grad_norm": 0.5173588991165161, "muon_lr": 0.00960698689956332, "adamw_lr": 0.0009606986899563319, "ngram_lr": 0.00048034934497816597, "weight_decay": 0.1, "tokens_per_second": 1165742.4702007251, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 25487.311614274513}
{"step": 560, "total_steps": 57221, "tokens": 293601280, "source": "stack_v3", "source_tokens": {"stack_v3": 220200960, "fineweb_edu": 73400320}, "loss": 2.4976490139961243, "ce_loss": 2.4865732192993164, "perplexity": 12.020015508120908, "z_loss": 110.75792503356934, "grad_norm": 0.42594170570373535, "muon_lr": 0.009781659388646289, "adamw_lr": 0.0009781659388646289, "ngram_lr": 0.0004890829694323144, "weight_decay": 0.1, "tokens_per_second": 1179802.053709451, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 25179.137997428654}
{"step": 570, "total_steps": 57221, "tokens": 298844160, "source": "stack_v3", "source_tokens": {"stack_v3": 224395264, "fineweb_edu": 74448896}, "loss": 2.2817051112651825, "ce_loss": 2.2703301906585693, "perplexity": 9.682597389512829, "z_loss": 113.74931526184082, "grad_norm": 0.39999920129776, "muon_lr": 0.009956331877729258, "adamw_lr": 0.0009956331877729258, "ngram_lr": 0.0004978165938864629, "weight_decay": 0.1, "tokens_per_second": 1117516.0972461472, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 26577.832671217388}
{"step": 580, "total_steps": 57221, "tokens": 304087040, "source": "stack_v3", "source_tokens": {"stack_v3": 228065280, "fineweb_edu": 76021760}, "loss": 2.27516970038414, "ce_loss": 2.263826072216034, "perplexity": 9.619824990111525, "z_loss": 113.43635368347168, "grad_norm": 0.457724928855896, "muon_lr": 0.010131004366812227, "adamw_lr": 0.0010131004366812227, "ngram_lr": 0.0005065502183406114, "weight_decay": 0.1, "tokens_per_second": 1123422.6681879798, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 26433.428664830048}
{"step": 590, "total_steps": 57221, "tokens": 309329920, "source": "stack_v3", "source_tokens": {"stack_v3": 232259584, "fineweb_edu": 77070336}, "loss": 1.778429165482521, "ce_loss": 1.76646126806736, "perplexity": 5.850114701042488, "z_loss": 119.67917823791504, "grad_norm": 0.40332913398742676, "muon_lr": 0.010305676855895196, "adamw_lr": 0.0010305676855895196, "ngram_lr": 0.0005152838427947598, "weight_decay": 0.1, "tokens_per_second": 1141544.2576554944, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 26009.21504434594}
{"step": 600, "total_steps": 57221, "tokens": 314572800, "source": "stack_v3", "source_tokens": {"stack_v3": 235929600, "fineweb_edu": 78643200}, "loss": 1.7942249476909637, "ce_loss": 1.7823019325733185, "perplexity": 5.943522271137271, "z_loss": 119.2305679321289, "grad_norm": 0.47754260897636414, "muon_lr": 0.010480349344978166, "adamw_lr": 0.0010480349344978166, "ngram_lr": 0.0005240174672489083, "weight_decay": 0.1, "tokens_per_second": 1162201.1039852982, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 25542.418690023478}
{"step": 610, "total_steps": 57221, "tokens": 319815680, "source": "stack_v3", "source_tokens": {"stack_v3": 240123904, "fineweb_edu": 79691776}, "loss": 2.0860775411129, "ce_loss": 2.074528783559799, "perplexity": 7.960794317909253, "z_loss": 115.48821067810059, "grad_norm": 0.40786662697792053, "muon_lr": 0.010655021834061137, "adamw_lr": 0.0010655021834061135, "ngram_lr": 0.0005327510917030568, "weight_decay": 0.1, "tokens_per_second": 414579.0029723624, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 71591.14211574919}
{"step": 620, "total_steps": 57221, "tokens": 325058560, "source": "stack_v3", "source_tokens": {"stack_v3": 243793920, "fineweb_edu": 81264640}, "loss": 2.2983479499816895, "ce_loss": 2.287291705608368, "perplexity": 9.84822962579682, "z_loss": 110.56234359741211, "grad_norm": 0.6163716316223145, "muon_lr": 0.010829694323144104, "adamw_lr": 0.0010829694323144104, "ngram_lr": 0.0005414847161572052, "weight_decay": 0.1, "tokens_per_second": 1079634.1593774538, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 27486.108310162555}
{"step": 630, "total_steps": 57221, "tokens": 330301440, "source": "stack_v3", "source_tokens": {"stack_v3": 247988224, "fineweb_edu": 82313216}, "loss": 1.8404811322689056, "ce_loss": 1.8286009728908539, "perplexity": 6.225171379971095, "z_loss": 118.8018684387207, "grad_norm": 0.7198107242584229, "muon_lr": 0.011004366812227074, "adamw_lr": 0.0011004366812227074, "ngram_lr": 0.0005502183406113537, "weight_decay": 0.1, "tokens_per_second": 1158053.5508800903, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 25620.31655397266}
{"step": 640, "total_steps": 57221, "tokens": 335544320, "source": "stack_v3", "source_tokens": {"stack_v3": 251658240, "fineweb_edu": 83886080}, "loss": 2.0630459934473038, "ce_loss": 2.051250323653221, "perplexity": 7.777619571143018, "z_loss": 117.9562816619873, "grad_norm": 0.47931626439094543, "muon_lr": 0.011179039301310045, "adamw_lr": 0.0011179039301310043, "ngram_lr": 0.0005589519650655021, "weight_decay": 0.1, "tokens_per_second": 1101941.2981442783, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 26920.17780797975}
{"step": 650, "total_steps": 57221, "tokens": 340787200, "source": "stack_v3", "source_tokens": {"stack_v3": 255852544, "fineweb_edu": 84934656}, "loss": 2.187092423439026, "ce_loss": 2.1760496497154236, "perplexity": 8.81142918396089, "z_loss": 110.42780303955078, "grad_norm": 1.1256159543991089, "muon_lr": 0.011353711790393014, "adamw_lr": 0.0011353711790393014, "ngram_lr": 0.0005676855895196507, "weight_decay": 0.1, "tokens_per_second": 1110947.499142675, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 26697.22270664293}
{"step": 660, "total_steps": 57221, "tokens": 346030080, "source": "stack_v3", "source_tokens": {"stack_v3": 259522560, "fineweb_edu": 86507520}, "loss": 1.865532010793686, "ce_loss": 1.8536615073680878, "perplexity": 6.383148732616881, "z_loss": 118.70465087890625, "grad_norm": 0.4679875075817108, "muon_lr": 0.011528384279475983, "adamw_lr": 0.0011528384279475984, "ngram_lr": 0.0005764192139737992, "weight_decay": 0.1, "tokens_per_second": 1167773.289941553, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 25393.60180218215}
{"step": 670, "total_steps": 57221, "tokens": 351272960, "source": "stack_v3", "source_tokens": {"stack_v3": 263716864, "fineweb_edu": 87556096}, "loss": 2.335373044013977, "ce_loss": 2.323859840631485, "perplexity": 10.21502668531166, "z_loss": 115.1320571899414, "grad_norm": 0.5141765475273132, "muon_lr": 0.011703056768558952, "adamw_lr": 0.0011703056768558953, "ngram_lr": 0.0005851528384279476, "weight_decay": 0.1, "tokens_per_second": 1155026.837703375, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 25669.297086596493}
{"step": 680, "total_steps": 57221, "tokens": 356515840, "source": "stack_v3", "source_tokens": {"stack_v3": 267386880, "fineweb_edu": 89128960}, "loss": 2.1287272572517395, "ce_loss": 2.1168857514858246, "perplexity": 8.305232613190352, "z_loss": 118.41510581970215, "grad_norm": 0.48233383893966675, "muon_lr": 0.011877729257641922, "adamw_lr": 0.0011877729257641922, "ngram_lr": 0.0005938864628820961, "weight_decay": 0.1, "tokens_per_second": 1127929.99314273, "data_wait_seconds": 1.6033649444580078e-05, "eta_seconds": 26281.315631482517}
{"step": 690, "total_steps": 57221, "tokens": 361758720, "source": "stack_v3", "source_tokens": {"stack_v3": 271581184, "fineweb_edu": 90177536}, "loss": 1.967759132385254, "ce_loss": 1.9555040299892426, "perplexity": 7.067480348162507, "z_loss": 122.55079650878906, "grad_norm": 0.39743971824645996, "muon_lr": 0.012052401746724891, "adamw_lr": 0.0012052401746724891, "ngram_lr": 0.0006026200873362446, "weight_decay": 0.1, "tokens_per_second": 1138975.4765230839, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 26021.84321867558}
{"step": 700, "total_steps": 57221, "tokens": 367001600, "source": "stack_v3", "source_tokens": {"stack_v3": 275251200, "fineweb_edu": 91750400}, "loss": 2.760310113430023, "ce_loss": 2.748952865600586, "perplexity": 15.626260519250483, "z_loss": 113.57221603393555, "grad_norm": 0.46410807967185974, "muon_lr": 0.01222707423580786, "adamw_lr": 0.001222707423580786, "ngram_lr": 0.000611353711790393, "weight_decay": 0.1, "tokens_per_second": 1170161.9704957341, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 25323.84332012269}
{"step": 710, "total_steps": 57221, "tokens": 372244480, "source": "stack_v3", "source_tokens": {"stack_v3": 279445504, "fineweb_edu": 92798976}, "loss": 2.2255484759807587, "ce_loss": 2.21407151222229, "perplexity": 9.15290680297771, "z_loss": 114.76944351196289, "grad_norm": 0.389450341463089, "muon_lr": 0.012401746724890831, "adamw_lr": 0.001240174672489083, "ngram_lr": 0.0006200873362445415, "weight_decay": 0.1, "tokens_per_second": 522712.25019943196, "data_wait_seconds": 2.199411392211914e-05, "eta_seconds": 56680.8134087082}
{"step": 720, "total_steps": 57221, "tokens": 377487360, "source": "stack_v3", "source_tokens": {"stack_v3": 283115520, "fineweb_edu": 94371840}, "loss": 2.0339884161949158, "ce_loss": 2.0216030180454254, "perplexity": 7.550418695232135, "z_loss": 123.8540153503418, "grad_norm": 0.5575564503669739, "muon_lr": 0.012576419213973799, "adamw_lr": 0.00125764192139738, "ngram_lr": 0.00062882096069869, "weight_decay": 0.1, "tokens_per_second": 1147031.658235543, "data_wait_seconds": 2.2411346435546875e-05, "eta_seconds": 25825.366220116146}
{"step": 730, "total_steps": 57221, "tokens": 382730240, "source": "stack_v3", "source_tokens": {"stack_v3": 287309824, "fineweb_edu": 95420416}, "loss": 2.0474328994750977, "ce_loss": 2.0348477363586426, "perplexity": 7.651087054406663, "z_loss": 125.85173034667969, "grad_norm": 0.35276174545288086, "muon_lr": 0.012751091703056768, "adamw_lr": 0.0012751091703056769, "ngram_lr": 0.0006375545851528384, "weight_decay": 0.1, "tokens_per_second": 1151111.405180177, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 25729.281828603005}
{"step": 740, "total_steps": 57221, "tokens": 387973120, "source": "stack_v3", "source_tokens": {"stack_v3": 290979840, "fineweb_edu": 96993280}, "loss": 2.018427014350891, "ce_loss": 2.005864918231964, "perplexity": 7.432519639231735, "z_loss": 125.62081336975098, "grad_norm": 0.47059616446495056, "muon_lr": 0.01292576419213974, "adamw_lr": 0.0012925764192139738, "ngram_lr": 0.0006462882096069869, "weight_decay": 0.1, "tokens_per_second": 1123981.3192865949, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 26345.657505050996}
{"step": 750, "total_steps": 57221, "tokens": 393216000, "source": "stack_v3", "source_tokens": {"stack_v3": 295174144, "fineweb_edu": 98041856}, "loss": 1.9763010144233704, "ce_loss": 1.9632561802864075, "perplexity": 7.122481430929393, "z_loss": 130.44859886169434, "grad_norm": 0.2927093207836151, "muon_lr": 0.013100436681222707, "adamw_lr": 0.0013100436681222707, "ngram_lr": 0.0006550218340611354, "weight_decay": 0.1, "tokens_per_second": 1163915.98782977, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 25437.217384739783}
{"step": 760, "total_steps": 57221, "tokens": 398458880, "source": "stack_v3", "source_tokens": {"stack_v3": 298844160, "fineweb_edu": 99614720}, "loss": 2.141558587551117, "ce_loss": 2.1287032067775726, "perplexity": 8.403961541659653, "z_loss": 128.5544776916504, "grad_norm": 0.3989570438861847, "muon_lr": 0.013275109170305678, "adamw_lr": 0.0013275109170305676, "ngram_lr": 0.0006637554585152838, "weight_decay": 0.1, "tokens_per_second": 1127884.0588235324, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 26245.19859858346}
{"step": 770, "total_steps": 57221, "tokens": 403701760, "source": "stack_v3", "source_tokens": {"stack_v3": 303038464, "fineweb_edu": 100663296}, "loss": 2.2895984053611755, "ce_loss": 2.276941418647766, "perplexity": 9.746823320585193, "z_loss": 126.56966972351074, "grad_norm": 0.2796696126461029, "muon_lr": 0.013449781659388647, "adamw_lr": 0.0013449781659388646, "ngram_lr": 0.0006724890829694323, "weight_decay": 0.1, "tokens_per_second": 1135295.4008693928, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 26069.248776429096}
{"step": 780, "total_steps": 57221, "tokens": 408944640, "source": "stack_v3", "source_tokens": {"stack_v3": 306708480, "fineweb_edu": 102236160}, "loss": 1.8199207782745361, "ce_loss": 1.8070180118083954, "perplexity": 6.092253290981699, "z_loss": 129.02774047851562, "grad_norm": 0.4615360498428345, "muon_lr": 0.013624454148471615, "adamw_lr": 0.0013624454148471617, "ngram_lr": 0.0006812227074235809, "weight_decay": 0.1, "tokens_per_second": 1136149.6124019313, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 26045.034066808876}
{"step": 790, "total_steps": 57221, "tokens": 414187520, "source": "stack_v3", "source_tokens": {"stack_v3": 310902784, "fineweb_edu": 103284736}, "loss": 1.8262935280799866, "ce_loss": 1.8136383891105652, "perplexity": 6.132720111395146, "z_loss": 126.55160903930664, "grad_norm": 0.2890819311141968, "muon_lr": 0.013799126637554586, "adamw_lr": 0.0013799126637554586, "ngram_lr": 0.0006899563318777293, "weight_decay": 0.1, "tokens_per_second": 467096.90861627075, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 63339.773683463456}
{"step": 800, "total_steps": 57221, "tokens": 419430400, "source": "stack_v3", "source_tokens": {"stack_v3": 314572800, "fineweb_edu": 104857600}, "loss": 2.023677349090576, "ce_loss": 2.011130154132843, "perplexity": 7.471756814314964, "z_loss": 125.47182273864746, "grad_norm": 0.4852356016635895, "muon_lr": 0.013973799126637555, "adamw_lr": 0.0013973799126637556, "ngram_lr": 0.0006986899563318778, "weight_decay": 0.1, "tokens_per_second": 1090722.0441829518, "data_wait_seconds": 2.473592758178711e-05, "eta_seconds": 27120.172144460954}
{"step": 810, "total_steps": 57221, "tokens": 424673280, "source": "stack_v3", "source_tokens": {"stack_v3": 318767104, "fineweb_edu": 105906176}, "loss": 2.1253366470336914, "ce_loss": 2.11316379904747, "perplexity": 8.274378387014208, "z_loss": 121.7285327911377, "grad_norm": 0.30445045232772827, "muon_lr": 0.014148471615720524, "adamw_lr": 0.0014148471615720525, "ngram_lr": 0.0007074235807860263, "weight_decay": 0.1, "tokens_per_second": 1114934.0647196409, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 26526.525339807384}
{"step": 820, "total_steps": 57221, "tokens": 429916160, "source": "stack_v3", "source_tokens": {"stack_v3": 322437120, "fineweb_edu": 107479040}, "loss": 1.8808155059814453, "ce_loss": 1.8680332899093628, "perplexity": 6.475548341263433, "z_loss": 127.82200050354004, "grad_norm": 0.34115463495254517, "muon_lr": 0.014323144104803492, "adamw_lr": 0.0014323144104803494, "ngram_lr": 0.0007161572052401747, "weight_decay": 0.1, "tokens_per_second": 1172681.5004341116, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 25215.784361784114}
{"step": 830, "total_steps": 57221, "tokens": 435159040, "source": "stack_v3", "source_tokens": {"stack_v3": 326631424, "fineweb_edu": 108527616}, "loss": 1.7582685351371765, "ce_loss": 1.7449378669261932, "perplexity": 5.72554571861457, "z_loss": 133.30664253234863, "grad_norm": 0.36833348870277405, "muon_lr": 0.014497816593886464, "adamw_lr": 0.0014497816593886464, "ngram_lr": 0.0007248908296943232, "weight_decay": 0.1, "tokens_per_second": 1141288.0192037455, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 25904.802698819894}
{"step": 840, "total_steps": 57221, "tokens": 440401920, "source": "stack_v3", "source_tokens": {"stack_v3": 330301440, "fineweb_edu": 110100480}, "loss": 1.818343311548233, "ce_loss": 1.8058218955993652, "perplexity": 6.0849706044088405, "z_loss": 125.21367263793945, "grad_norm": 0.42261672019958496, "muon_lr": 0.014672489082969434, "adamw_lr": 0.001467248908296943, "ngram_lr": 0.0007336244541484715, "weight_decay": 0.1, "tokens_per_second": 439980.27363642474, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 67183.91675992822}
{"step": 850, "total_steps": 57221, "tokens": 445644800, "source": "stack_v3", "source_tokens": {"stack_v3": 334495744, "fineweb_edu": 111149056}, "loss": 1.599895179271698, "ce_loss": 1.587097555398941, "perplexity": 4.889536703799026, "z_loss": 127.9763412475586, "grad_norm": 0.4377575218677521, "muon_lr": 0.014847161572052401, "adamw_lr": 0.0014847161572052402, "ngram_lr": 0.0007423580786026201, "weight_decay": 0.1, "tokens_per_second": 1132466.4000164797, "data_wait_seconds": 1.8596649169921875e-05, "eta_seconds": 26097.335161175575}
{"step": 860, "total_steps": 57221, "tokens": 450887680, "source": "stack_v3", "source_tokens": {"stack_v3": 338165760, "fineweb_edu": 112721920}, "loss": 1.5765438675880432, "ce_loss": 1.5630785524845123, "perplexity": 4.773494100093984, "z_loss": 134.65293312072754, "grad_norm": 0.5567122101783752, "muon_lr": 0.01502183406113537, "adamw_lr": 0.0015021834061135371, "ngram_lr": 0.0007510917030567686, "weight_decay": 0.1, "tokens_per_second": 1177214.0533952736, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 25100.883084750505}
{"step": 870, "total_steps": 57221, "tokens": 456130560, "source": "stack_v3", "source_tokens": {"stack_v3": 342360064, "fineweb_edu": 113770496}, "loss": 1.4075402319431305, "ce_loss": 1.3939166218042374, "perplexity": 4.030605536246466, "z_loss": 136.23588752746582, "grad_norm": 0.32304149866104126, "muon_lr": 0.015196506550218342, "adamw_lr": 0.001519650655021834, "ngram_lr": 0.000759825327510917, "weight_decay": 0.1, "tokens_per_second": 1129945.9304505337, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 26146.268280483317}
{"step": 880, "total_steps": 57221, "tokens": 461373440, "source": "stack_v3", "source_tokens": {"stack_v3": 346030080, "fineweb_edu": 115343360}, "loss": 1.6886975765228271, "ce_loss": 1.6762515902519226, "perplexity": 5.3454813163368415, "z_loss": 124.45956802368164, "grad_norm": 0.4644430875778198, "muon_lr": 0.015371179039301311, "adamw_lr": 0.001537117903930131, "ngram_lr": 0.0007685589519650655, "weight_decay": 0.1, "tokens_per_second": 1142276.4000981548, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 25859.43871156033}
{"step": 890, "total_steps": 57221, "tokens": 466616320, "source": "stack_v3", "source_tokens": {"stack_v3": 350224384, "fineweb_edu": 116391936}, "loss": 1.0925141870975494, "ce_loss": 1.0782178416848183, "perplexity": 2.939436339422403, "z_loss": 142.9632625579834, "grad_norm": 0.2588732838630676, "muon_lr": 0.01554585152838428, "adamw_lr": 0.001554585152838428, "ngram_lr": 0.000777292576419214, "weight_decay": 0.1, "tokens_per_second": 1144728.6379195498, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 25799.46259899158}
{"step": 900, "total_steps": 57221, "tokens": 471859200, "source": "stack_v3", "source_tokens": {"stack_v3": 353894400, "fineweb_edu": 117964800}, "loss": 1.2785935550928116, "ce_loss": 1.2645284086465836, "perplexity": 3.5414222381374394, "z_loss": 140.65145874023438, "grad_norm": 0.4218919575214386, "muon_lr": 0.01572052401746725, "adamw_lr": 0.0015720524017467249, "ngram_lr": 0.0007860262008733624, "weight_decay": 0.1, "tokens_per_second": 1186509.8523800953, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 24886.5533992555}
{"step": 910, "total_steps": 57221, "tokens": 477102080, "source": "stack_v3", "source_tokens": {"stack_v3": 358088704, "fineweb_edu": 119013376}, "loss": 1.1158397942781448, "ce_loss": 1.1014912128448486, "perplexity": 3.0086492167759875, "z_loss": 143.48575019836426, "grad_norm": 0.23971810936927795, "muon_lr": 0.01589519650655022, "adamw_lr": 0.0015895196506550218, "ngram_lr": 0.0007947598253275109, "weight_decay": 0.1, "tokens_per_second": 1177096.8471491942, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 25081.112052505603}
{"step": 920, "total_steps": 57221, "tokens": 482344960, "source": "stack_v3", "source_tokens": {"stack_v3": 361758720, "fineweb_edu": 120586240}, "loss": 1.9349961876869202, "ce_loss": 1.9217028617858887, "perplexity": 6.83258351392424, "z_loss": 132.93339347839355, "grad_norm": 0.4357107877731323, "muon_lr": 0.016069868995633188, "adamw_lr": 0.001606986899563319, "ngram_lr": 0.0008034934497816595, "weight_decay": 0.1, "tokens_per_second": 1030366.5613517292, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 28647.72222545347}
{"step": 930, "total_steps": 57221, "tokens": 487587840, "source": "stack_v3", "source_tokens": {"stack_v3": 365953024, "fineweb_edu": 121634816}, "loss": 1.800808146595955, "ce_loss": 1.7870593070983887, "perplexity": 5.971865198082287, "z_loss": 137.48793601989746, "grad_norm": 0.2445349097251892, "muon_lr": 0.016244541484716157, "adamw_lr": 0.0016244541484716159, "ngram_lr": 0.0008122270742358079, "weight_decay": 0.1, "tokens_per_second": 1180370.234883678, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 25002.67398127704}
{"step": 940, "total_steps": 57221, "tokens": 492830720, "source": "stack_v3", "source_tokens": {"stack_v3": 369623040, "fineweb_edu": 123207680}, "loss": 1.67823426425457, "ce_loss": 1.6644434928894043, "perplexity": 5.282732553348112, "z_loss": 137.90752792358398, "grad_norm": 0.3755902051925659, "muon_lr": 0.016419213973799127, "adamw_lr": 0.0016419213973799128, "ngram_lr": 0.0008209606986899564, "weight_decay": 0.1, "tokens_per_second": 456321.92595614877, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 64663.05386963929}
{"step": 950, "total_steps": 57221, "tokens": 498073600, "source": "stack_v3", "source_tokens": {"stack_v3": 373817344, "fineweb_edu": 124256256}, "loss": 1.218914434313774, "ce_loss": 1.2048529833555222, "perplexity": 3.3362685550027376, "z_loss": 140.61444091796875, "grad_norm": 0.2113986760377884, "muon_lr": 0.016593886462882096, "adamw_lr": 0.0016593886462882097, "ngram_lr": 0.0008296943231441049, "weight_decay": 0.1, "tokens_per_second": 1142767.432060882, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 25816.212093825456}
{"step": 960, "total_steps": 57221, "tokens": 503316480, "source": "stack_v3", "source_tokens": {"stack_v3": 377487360, "fineweb_edu": 125829120}, "loss": 1.160791039466858, "ce_loss": 1.1466550827026367, "perplexity": 3.1476466635912645, "z_loss": 141.3596363067627, "grad_norm": 0.2508845329284668, "muon_lr": 0.016768558951965065, "adamw_lr": 0.0016768558951965064, "ngram_lr": 0.0008384279475982532, "weight_decay": 0.1, "tokens_per_second": 1138573.1109216327, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 25906.7101067612}
{"step": 970, "total_steps": 57221, "tokens": 508559360, "source": "stack_v3", "source_tokens": {"stack_v3": 381681664, "fineweb_edu": 126877696}, "loss": 0.9291881993412971, "ce_loss": 0.9146254286170006, "perplexity": 2.4958402064773053, "z_loss": 145.62752532958984, "grad_norm": 0.2333381623029709, "muon_lr": 0.016943231441048038, "adamw_lr": 0.0016943231441048034, "ngram_lr": 0.0008471615720524017, "weight_decay": 0.1, "tokens_per_second": 1144227.5245323377, "data_wait_seconds": 1.7762184143066406e-05, "eta_seconds": 25774.10524366959}
{"step": 980, "total_steps": 57221, "tokens": 513802240, "source": "stack_v3", "source_tokens": {"stack_v3": 385351680, "fineweb_edu": 128450560}, "loss": 1.2829001545906067, "ce_loss": 1.2694523185491562, "perplexity": 3.5589028834081606, "z_loss": 134.4783935546875, "grad_norm": 0.4740423560142517, "muon_lr": 0.017117903930131007, "adamw_lr": 0.0017117903930131005, "ngram_lr": 0.0008558951965065503, "weight_decay": 0.1, "tokens_per_second": 1175040.7188458669, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 25093.766783640953}
{"step": 990, "total_steps": 57221, "tokens": 519045120, "source": "stack_v3", "source_tokens": {"stack_v3": 389545984, "fineweb_edu": 129499136}, "loss": 0.7919897884130478, "ce_loss": 0.7773152887821198, "perplexity": 2.175623497075622, "z_loss": 146.7449722290039, "grad_norm": 0.2650759220123291, "muon_lr": 0.017292576419213973, "adamw_lr": 0.0017292576419213974, "ngram_lr": 0.0008646288209606987, "weight_decay": 0.1, "tokens_per_second": 1107699.465298621, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 26614.578957165362}
{"step": 1000, "total_steps": 57221, "tokens": 524288000, "source": "stack_v3", "source_tokens": {"stack_v3": 393216000, "fineweb_edu": 131072000}, "loss": 1.2355122417211533, "ce_loss": 1.2216764390468597, "perplexity": 3.3928709098152403, "z_loss": 138.35804176330566, "grad_norm": 0.2927790582180023, "muon_lr": 0.017467248908296942, "adamw_lr": 0.0017467248908296944, "ngram_lr": 0.0008733624454148472, "weight_decay": 0.1, "tokens_per_second": 456230.99250623106, "data_wait_seconds": 2.2530555725097656e-05, "eta_seconds": 64606.991818069946}
{"step": 1010, "total_steps": 57221, "tokens": 529530880, "source": "stack_v3", "source_tokens": {"stack_v3": 397410304, "fineweb_edu": 132120576}, "loss": 1.009154051542282, "ce_loss": 0.9948360472917557, "perplexity": 2.704280930748817, "z_loss": 143.18000411987305, "grad_norm": 0.1754804104566574, "muon_lr": 0.01764192139737991, "adamw_lr": 0.0017641921397379913, "ngram_lr": 0.0008820960698689956, "weight_decay": 0.1, "tokens_per_second": 1151306.3684170737, "data_wait_seconds": 1.5676021575927734e-05, "eta_seconds": 25597.41692432295}
{"step": 1020, "total_steps": 57221, "tokens": 534773760, "source": "stack_v3", "source_tokens": {"stack_v3": 401080320, "fineweb_edu": 133693440}, "loss": 1.0604883655905724, "ce_loss": 1.0465775281190872, "perplexity": 2.8478876046714108, "z_loss": 139.1082649230957, "grad_norm": 0.2141922116279602, "muon_lr": 0.017816593886462884, "adamw_lr": 0.0017816593886462882, "ngram_lr": 0.0008908296943231441, "weight_decay": 0.1, "tokens_per_second": 1190266.7281867662, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 24755.145667968784}
{"step": 1030, "total_steps": 57221, "tokens": 540016640, "source": "stack_v3", "source_tokens": {"stack_v3": 405274624, "fineweb_edu": 134742016}, "loss": 1.1805690675973892, "ce_loss": 1.166677013039589, "perplexity": 3.211303768328034, "z_loss": 138.92086219787598, "grad_norm": 0.2695462107658386, "muon_lr": 0.017991266375545854, "adamw_lr": 0.0017991266375545851, "ngram_lr": 0.0008995633187772926, "weight_decay": 0.1, "tokens_per_second": 1160751.789474473, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 25380.088686607083}
{"step": 1040, "total_steps": 57221, "tokens": 545259520, "source": "stack_v3", "source_tokens": {"stack_v3": 408944640, "fineweb_edu": 136314880}, "loss": 1.151433065533638, "ce_loss": 1.1373330056667328, "perplexity": 3.1184404019264775, "z_loss": 141.0005111694336, "grad_norm": 0.2372133731842041, "muon_lr": 0.018165938864628823, "adamw_lr": 0.001816593886462882, "ngram_lr": 0.000908296943231441, "weight_decay": 0.1, "tokens_per_second": 1151141.775326618, "data_wait_seconds": 2.4199485778808594e-05, "eta_seconds": 25587.413393665338}
{"step": 1050, "total_steps": 57221, "tokens": 550502400, "source": "stack_v3", "source_tokens": {"stack_v3": 413138944, "fineweb_edu": 137363456}, "loss": 1.0938395261764526, "ce_loss": 1.0800335109233856, "perplexity": 2.944778231649784, "z_loss": 138.06000328063965, "grad_norm": 0.17562632262706757, "muon_lr": 0.01834061135371179, "adamw_lr": 0.0018340611353711792, "ngram_lr": 0.0009170305676855896, "weight_decay": 0.1, "tokens_per_second": 1180184.7495831696, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 24953.29448241158}
{"step": 1060, "total_steps": 57221, "tokens": 555745280, "source": "stack_v3", "source_tokens": {"stack_v3": 416808960, "fineweb_edu": 138936320}, "loss": 1.1054641902446747, "ce_loss": 1.0914408564567566, "perplexity": 2.9785626634446625, "z_loss": 140.2330837249756, "grad_norm": 0.3460935056209564, "muon_lr": 0.018515283842794758, "adamw_lr": 0.0018515283842794761, "ngram_lr": 0.0009257641921397381, "weight_decay": 0.1, "tokens_per_second": 1176391.0787338614, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 25029.30806963495}
{"step": 1070, "total_steps": 57221, "tokens": 560988160, "source": "stack_v3", "source_tokens": {"stack_v3": 421003264, "fineweb_edu": 139984896}, "loss": 0.9406368881464005, "ce_loss": 0.9268456101417542, "perplexity": 2.5265269437298747, "z_loss": 137.91271591186523, "grad_norm": 0.1718050092458725, "muon_lr": 0.01868995633187773, "adamw_lr": 0.001868995633187773, "ngram_lr": 0.0009344978165938865, "weight_decay": 0.1, "tokens_per_second": 1148040.4454454235, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 25642.835108111612}
{"step": 1080, "total_steps": 57221, "tokens": 566231040, "source": "stack_v3", "source_tokens": {"stack_v3": 424673280, "fineweb_edu": 141557760}, "loss": 1.0371598601341248, "ce_loss": 1.0234312415122986, "perplexity": 2.782726608638805, "z_loss": 137.2860927581787, "grad_norm": 0.2589509189128876, "muon_lr": 0.0188646288209607, "adamw_lr": 0.00188646288209607, "ngram_lr": 0.000943231441048035, "weight_decay": 0.1, "tokens_per_second": 1126093.4579844328, "data_wait_seconds": 2.1636486053466797e-05, "eta_seconds": 26137.945080226986}
{"step": 1090, "total_steps": 57221, "tokens": 571473920, "source": "stack_v3", "source_tokens": {"stack_v3": 428867584, "fineweb_edu": 142606336}, "loss": 1.1034408658742905, "ce_loss": 1.0896879732608795, "perplexity": 2.9733461642965766, "z_loss": 137.5290470123291, "grad_norm": 0.1588793694972992, "muon_lr": 0.01903930131004367, "adamw_lr": 0.001903930131004367, "ngram_lr": 0.0009519650655021835, "weight_decay": 0.1, "tokens_per_second": 1175131.2659856107, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 25042.75644076034}
{"step": 1100, "total_steps": 57221, "tokens": 576716800, "source": "stack_v3", "source_tokens": {"stack_v3": 432537600, "fineweb_edu": 144179200}, "loss": 1.0486771911382675, "ce_loss": 1.0354757755994797, "perplexity": 2.8164459131485846, "z_loss": 132.01401329040527, "grad_norm": 0.1829717457294464, "muon_lr": 0.01921397379912664, "adamw_lr": 0.0019213973799126639, "ngram_lr": 0.0009606986899563319, "weight_decay": 0.1, "tokens_per_second": 908364.1507663658, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 32391.506396610057}
{"step": 1110, "total_steps": 57221, "tokens": 581959680, "source": "stack_v3", "source_tokens": {"stack_v3": 436731904, "fineweb_edu": 145227776}, "loss": 1.1343805640935898, "ce_loss": 1.1213190406560898, "perplexity": 3.0688995379860677, "z_loss": 130.61516189575195, "grad_norm": 0.22276577353477478, "muon_lr": 0.019388646288209608, "adamw_lr": 0.0019388646288209608, "ngram_lr": 0.0009694323144104804, "weight_decay": 0.1, "tokens_per_second": 1168050.8049197844, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 25185.582849729104}
{"step": 1120, "total_steps": 57221, "tokens": 587202560, "source": "stack_v3", "source_tokens": {"stack_v3": 440401920, "fineweb_edu": 146800640}, "loss": 1.0763913094997406, "ce_loss": 1.0629292279481888, "perplexity": 2.8948382230130916, "z_loss": 134.62096214294434, "grad_norm": 0.4199754297733307, "muon_lr": 0.019563318777292577, "adamw_lr": 0.0019563318777292577, "ngram_lr": 0.0009781659388646289, "weight_decay": 0.1, "tokens_per_second": 1124826.6386576372, "data_wait_seconds": 2.205371856689453e-05, "eta_seconds": 26148.738329224754}
{"step": 1130, "total_steps": 57221, "tokens": 592445440, "source": "stack_v3", "source_tokens": {"stack_v3": 444596224, "fineweb_edu": 147849216}, "loss": 0.9773968756198883, "ce_loss": 0.9643222466111183, "perplexity": 2.6230093004567876, "z_loss": 130.7464714050293, "grad_norm": 0.2572457790374756, "muon_lr": 0.019737991266375546, "adamw_lr": 0.0019737991266375546, "ngram_lr": 0.0009868995633187773, "weight_decay": 0.1, "tokens_per_second": 1178844.819105254, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 24946.077790222127}
{"step": 1140, "total_steps": 57221, "tokens": 597688320, "source": "stack_v3", "source_tokens": {"stack_v3": 448266240, "fineweb_edu": 149422080}, "loss": 0.9597291499376297, "ce_loss": 0.9462071657180786, "perplexity": 2.575921066067476, "z_loss": 135.21975898742676, "grad_norm": 0.17807848751544952, "muon_lr": 0.019912663755458516, "adamw_lr": 0.0019912663755458516, "ngram_lr": 0.0009956331877729258, "weight_decay": 0.1, "tokens_per_second": 1146683.312545855, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 25641.178657009732}
{"step": 1150, "total_steps": 57221, "tokens": 602931200, "source": "stack_v3", "source_tokens": {"stack_v3": 452460544, "fineweb_edu": 150470656}, "loss": 0.9840647354722023, "ce_loss": 0.9711278676986694, "perplexity": 2.6409213901772244, "z_loss": 129.36869430541992, "grad_norm": 0.1735108494758606, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 428059.01980932633, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 68675.2700903127}
{"step": 1160, "total_steps": 57221, "tokens": 608174080, "source": "stack_v3", "source_tokens": {"stack_v3": 456130560, "fineweb_edu": 152043520}, "loss": 1.3216141909360886, "ce_loss": 1.3095000237226486, "perplexity": 3.704321176424914, "z_loss": 121.14167785644531, "grad_norm": 0.1918349713087082, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132684.6188422644, "data_wait_seconds": 2.199411392211914e-05, "eta_seconds": 25948.817023790674}
{"step": 1170, "total_steps": 57221, "tokens": 613416960, "source": "stack_v3", "source_tokens": {"stack_v3": 460324864, "fineweb_edu": 153092096}, "loss": 1.2273147702217102, "ce_loss": 1.2150686085224152, "perplexity": 3.370525303149851, "z_loss": 122.46148300170898, "grad_norm": 0.17451494932174683, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1158261.6102131153, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 25371.28294754843}
{"step": 1180, "total_steps": 57221, "tokens": 618659840, "source": "stack_v3", "source_tokens": {"stack_v3": 463994880, "fineweb_edu": 154664960}, "loss": 1.0606137067079544, "ce_loss": 1.0474499017000198, "perplexity": 2.8503731105664007, "z_loss": 131.6380958557129, "grad_norm": 0.16353186964988708, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155627.938169777, "data_wait_seconds": 2.3365020751953125e-05, "eta_seconds": 25424.56718944735}
{"step": 1190, "total_steps": 57221, "tokens": 623902720, "source": "stack_v3", "source_tokens": {"stack_v3": 468189184, "fineweb_edu": 155713536}, "loss": 1.3061764985322952, "ce_loss": 1.2932724356651306, "perplexity": 3.6446940887331505, "z_loss": 129.0405044555664, "grad_norm": 0.17040258646011353, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164196.9111847857, "data_wait_seconds": 1.728534698486328e-05, "eta_seconds": 25232.92837987724}
{"step": 1200, "total_steps": 57221, "tokens": 629145600, "source": "stack_v3", "source_tokens": {"stack_v3": 471859200, "fineweb_edu": 157286400}, "loss": 0.960249587893486, "ce_loss": 0.9464428424835205, "perplexity": 2.5765282223558654, "z_loss": 138.06732368469238, "grad_norm": 0.17849119007587433, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1072282.4099426395, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 27390.969139903318}
{"step": 1210, "total_steps": 57221, "tokens": 634388480, "source": "stack_v3", "source_tokens": {"stack_v3": 476053504, "fineweb_edu": 158334976}, "loss": 0.7912009116262197, "ce_loss": 0.777562215924263, "perplexity": 2.176160783900755, "z_loss": 136.38709259033203, "grad_norm": 0.1531347930431366, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148102.3622162517, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 25577.520338268252}
{"step": 1220, "total_steps": 57221, "tokens": 639631360, "source": "stack_v3", "source_tokens": {"stack_v3": 479723520, "fineweb_edu": 159907840}, "loss": 0.8394445488229394, "ce_loss": 0.8263269616290927, "perplexity": 2.2849107434347964, "z_loss": 131.17584037780762, "grad_norm": 0.12707361578941345, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1176008.134506099, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 24966.12717082165}
{"step": 1230, "total_steps": 57221, "tokens": 644874240, "source": "stack_v3", "source_tokens": {"stack_v3": 483917824, "fineweb_edu": 160956416}, "loss": 1.2336734384298325, "ce_loss": 1.2216277495026588, "perplexity": 3.392705716498735, "z_loss": 120.45683860778809, "grad_norm": 0.15579822659492493, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164379.8707864415, "data_wait_seconds": 2.300739288330078e-05, "eta_seconds": 25210.95262508537}
{"step": 1240, "total_steps": 57221, "tokens": 650117120, "source": "stack_v3", "source_tokens": {"stack_v3": 487587840, "fineweb_edu": 162529280}, "loss": 1.2586051896214485, "ce_loss": 1.2469759956002235, "perplexity": 3.479804087826967, "z_loss": 116.2917594909668, "grad_norm": 0.18673951923847198, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134602.6760716788, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 25867.983126584673}
{"step": 1250, "total_steps": 57221, "tokens": 655360000, "source": "stack_v3", "source_tokens": {"stack_v3": 491782144, "fineweb_edu": 163577856}, "loss": 1.4030476808547974, "ce_loss": 1.39150071144104, "perplexity": 4.020879707657026, "z_loss": 115.46968650817871, "grad_norm": 0.15824687480926514, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165556.1785787032, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 25176.512757869205}
{"step": 1260, "total_steps": 57221, "tokens": 660602880, "source": "stack_v3", "source_tokens": {"stack_v3": 495452160, "fineweb_edu": 165150720}, "loss": 0.9676641821861267, "ce_loss": 0.9545419812202454, "perplexity": 2.5974806155246584, "z_loss": 131.2219352722168, "grad_norm": 0.19038210809230804, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155295.7168781036, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 25395.57335093594}
{"step": 1270, "total_steps": 57221, "tokens": 665845760, "source": "stack_v3", "source_tokens": {"stack_v3": 499646464, "fineweb_edu": 166199296}, "loss": 1.0908551961183548, "ce_loss": 1.0790251642465591, "perplexity": 2.9418103708736525, "z_loss": 118.3004322052002, "grad_norm": 0.13897517323493958, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141413.3098762142, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 25699.8529684057}
{"step": 1280, "total_steps": 57221, "tokens": 671088640, "source": "stack_v3", "source_tokens": {"stack_v3": 503316480, "fineweb_edu": 167772160}, "loss": 1.017775945365429, "ce_loss": 1.00498366355896, "perplexity": 2.7318626435483804, "z_loss": 127.9229621887207, "grad_norm": 0.38077500462532043, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154416.4838389177, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 25405.83209836809}
{"step": 1290, "total_steps": 57221, "tokens": 676331520, "source": "stack_v3", "source_tokens": {"stack_v3": 507510784, "fineweb_edu": 168820736}, "loss": 1.2217202335596085, "ce_loss": 1.2100280225276947, "perplexity": 3.353578626982268, "z_loss": 116.92229461669922, "grad_norm": 0.13723576068878174, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1092143.9699475355, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 26849.636391261356}
{"step": 1300, "total_steps": 57221, "tokens": 681574400, "source": "stack_v3", "source_tokens": {"stack_v3": 511180800, "fineweb_edu": 170393600}, "loss": 1.2474355697631836, "ce_loss": 1.2357494980096817, "perplexity": 3.440956544554114, "z_loss": 116.86080169677734, "grad_norm": 0.1732737123966217, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1176449.2310559838, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 24921.114167998312}
{"step": 1310, "total_steps": 57221, "tokens": 686817280, "source": "stack_v3", "source_tokens": {"stack_v3": 515375104, "fineweb_edu": 171442176}, "loss": 0.9309705346822739, "ce_loss": 0.9194552302360535, "perplexity": 2.507923776681487, "z_loss": 115.15304565429688, "grad_norm": 0.10650704801082611, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1133247.2656264333, "data_wait_seconds": 2.3126602172851562e-05, "eta_seconds": 25866.537347254343}
{"step": 1320, "total_steps": 57221, "tokens": 692060160, "source": "stack_v3", "source_tokens": {"stack_v3": 519045120, "fineweb_edu": 173015040}, "loss": 0.8398079723119736, "ce_loss": 0.8279343321919441, "perplexity": 2.2885863949782013, "z_loss": 118.73644065856934, "grad_norm": 0.298076331615448, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 483163.5871451897, "data_wait_seconds": 2.384185791015625e-05, "eta_seconds": 60658.420087424805}
{"step": 1330, "total_steps": 57221, "tokens": 697303040, "source": "stack_v3", "source_tokens": {"stack_v3": 523239424, "fineweb_edu": 174063616}, "loss": 0.5931516773998737, "ce_loss": 0.5811739675700665, "perplexity": 1.7881364131873851, "z_loss": 119.77708435058594, "grad_norm": 0.17166303098201752, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1139400.1469090432, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 25717.652432722738}
{"step": 1340, "total_steps": 57221, "tokens": 702545920, "source": "stack_v3", "source_tokens": {"stack_v3": 526909440, "fineweb_edu": 175636480}, "loss": 0.7679371777921915, "ce_loss": 0.7554420456290245, "perplexity": 2.1285522326323534, "z_loss": 124.95145606994629, "grad_norm": 0.22804395854473114, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1181598.8349161074, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 24794.755389277358}
{"step": 1350, "total_steps": 57221, "tokens": 707788800, "source": "stack_v3", "source_tokens": {"stack_v3": 531103744, "fineweb_edu": 176685056}, "loss": 0.9154481459408998, "ce_loss": 0.9043449368327856, "perplexity": 2.470313181749604, "z_loss": 111.0320053100586, "grad_norm": 0.15166400372982025, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137174.3886586179, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 25758.767953393984}
{"step": 1360, "total_steps": 57221, "tokens": 713031680, "source": "stack_v3", "source_tokens": {"stack_v3": 534773760, "fineweb_edu": 178257920}, "loss": 0.9065326899290085, "ce_loss": 0.8948780279606581, "perplexity": 2.447037301129101, "z_loss": 116.54674530029297, "grad_norm": 0.2202533781528473, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 748120.2293557912, "data_wait_seconds": 2.1159648895263672e-05, "eta_seconds": 39147.40862604277}
{"step": 1370, "total_steps": 57221, "tokens": 718274560, "source": "stack_v3", "source_tokens": {"stack_v3": 538968064, "fineweb_edu": 179306496}, "loss": 1.156113924458623, "ce_loss": 1.1460130829364061, "perplexity": 3.145626523703116, "z_loss": 101.00873565673828, "grad_norm": 0.14351440966129303, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1117186.8645981806, "data_wait_seconds": 2.378225326538086e-05, "eta_seconds": 26210.23068556376}
{"step": 1380, "total_steps": 57221, "tokens": 723517440, "source": "stack_v3", "source_tokens": {"stack_v3": 542638080, "fineweb_edu": 180879360}, "loss": 0.799424447119236, "ce_loss": 0.7885480243712664, "perplexity": 2.2001994698576235, "z_loss": 108.76421928405762, "grad_norm": 0.26397940516471863, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1182443.9290384133, "data_wait_seconds": 2.288818359375e-05, "eta_seconds": 24759.298805659404}
{"step": 1390, "total_steps": 57221, "tokens": 728760320, "source": "stack_v3", "source_tokens": {"stack_v3": 546832384, "fineweb_edu": 181927936}, "loss": 0.7764668017625809, "ce_loss": 0.7660478614270687, "perplexity": 2.1512474031921034, "z_loss": 104.18937873840332, "grad_norm": 0.11468910425901413, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129806.8331022195, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 25908.180781335104}
{"step": 1400, "total_steps": 57221, "tokens": 734003200, "source": "stack_v3", "source_tokens": {"stack_v3": 550502400, "fineweb_edu": 183500800}, "loss": 0.6812927238643169, "ce_loss": 0.6710645817220211, "perplexity": 1.956318873763095, "z_loss": 102.28143501281738, "grad_norm": 0.14548316597938538, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131839.2778210898, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 25857.025262756506}
{"step": 1410, "total_steps": 57221, "tokens": 739246080, "source": "stack_v3", "source_tokens": {"stack_v3": 554696704, "fineweb_edu": 184549376}, "loss": 0.5786084719002247, "ce_loss": 0.5681641846895218, "perplexity": 1.7650238176077226, "z_loss": 104.44287109375, "grad_norm": 0.12058164924383163, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146750.2258677727, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 25516.239944803765}
{"step": 1420, "total_steps": 57221, "tokens": 744488960, "source": "stack_v3", "source_tokens": {"stack_v3": 558366720, "fineweb_edu": 186122240}, "loss": 0.8998972475528717, "ce_loss": 0.8900757171213627, "perplexity": 2.4353140392777908, "z_loss": 98.21522331237793, "grad_norm": 0.1609669327735901, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1178341.306058834, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24827.705597328255}
{"step": 1430, "total_steps": 57221, "tokens": 749731840, "source": "stack_v3", "source_tokens": {"stack_v3": 562561024, "fineweb_edu": 187170816}, "loss": 0.8218212611973286, "ce_loss": 0.8116431627422571, "perplexity": 2.2516047016486, "z_loss": 101.78091430664062, "grad_norm": 0.24722826480865479, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151715.0109783623, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 25397.140682530826}
{"step": 1440, "total_steps": 57221, "tokens": 754974720, "source": "stack_v3", "source_tokens": {"stack_v3": 566231040, "fineweb_edu": 188743680}, "loss": 0.7864389456808567, "ce_loss": 0.7771915346384048, "perplexity": 2.175354271311941, "z_loss": 92.47417259216309, "grad_norm": 0.14117446541786194, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 723580.2250978666, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 40417.115152704064}
{"step": 1450, "total_steps": 57221, "tokens": 760217600, "source": "stack_v3", "source_tokens": {"stack_v3": 570425344, "fineweb_edu": 189792256}, "loss": 0.9643470887094736, "ce_loss": 0.9550052657723427, "perplexity": 2.5986842669632333, "z_loss": 93.41815948486328, "grad_norm": 0.15515325963497162, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1109060.2483431476, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 26364.467073526466}
{"step": 1460, "total_steps": 57221, "tokens": 765460480, "source": "stack_v3", "source_tokens": {"stack_v3": 574095360, "fineweb_edu": 191365120}, "loss": 0.5090117137879133, "ce_loss": 0.4985832367092371, "perplexity": 1.6463870768176396, "z_loss": 104.28475189208984, "grad_norm": 0.12718352675437927, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 475289.0161204269, "data_wait_seconds": 1.5718190670013428, "eta_seconds": 61508.97354756598}
{"step": 1470, "total_steps": 57221, "tokens": 770703360, "source": "stack_v3", "source_tokens": {"stack_v3": 578289664, "fineweb_edu": 192413696}, "loss": 0.6543395593762398, "ce_loss": 0.6449245363473892, "perplexity": 1.9058432019558007, "z_loss": 94.15007591247559, "grad_norm": 0.16823658347129822, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1156458.4156634668, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 25274.83586448803}
{"step": 1480, "total_steps": 57221, "tokens": 775946240, "source": "stack_v3", "source_tokens": {"stack_v3": 581959680, "fineweb_edu": 193986560}, "loss": 0.9212051760405302, "ce_loss": 0.9102451521903276, "perplexity": 2.4849316451548478, "z_loss": 109.60044479370117, "grad_norm": 0.1297895461320877, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130922.4328882548, "data_wait_seconds": 1.7821788787841797e-05, "eta_seconds": 25840.900233418215}
{"step": 1490, "total_steps": 57221, "tokens": 781189120, "source": "stack_v3", "source_tokens": {"stack_v3": 586153984, "fineweb_edu": 195035136}, "loss": 0.6576512716710567, "ce_loss": 0.647725434973836, "perplexity": 1.9111887582483484, "z_loss": 99.25836753845215, "grad_norm": 0.13805502653121948, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1172438.4107405525, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 24921.403642468857}
{"step": 1500, "total_steps": 57221, "tokens": 786432000, "source": "stack_v3", "source_tokens": {"stack_v3": 589824000, "fineweb_edu": 196608000}, "loss": 0.8708221688866615, "ce_loss": 0.8618838312104344, "perplexity": 2.3676166854262153, "z_loss": 89.38346099853516, "grad_norm": 0.1861732304096222, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1171701.3176582418, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 24932.606594986286}
{"step": 1510, "total_steps": 57221, "tokens": 791674880, "source": "stack_v3", "source_tokens": {"stack_v3": 594018304, "fineweb_edu": 197656576}, "loss": 0.6317830868065357, "ce_loss": 0.6230982188135386, "perplexity": 1.8646963387875206, "z_loss": 86.84858131408691, "grad_norm": 0.11871277540922165, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108762.3089098118, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 26343.180035330588}
{"step": 1520, "total_steps": 57221, "tokens": 796917760, "source": "stack_v3", "source_tokens": {"stack_v3": 597688320, "fineweb_edu": 199229440}, "loss": 0.6767870113253593, "ce_loss": 0.6677466705441475, "perplexity": 1.949838737707011, "z_loss": 90.40356254577637, "grad_norm": 0.15012064576148987, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 747850.6172633861, "data_wait_seconds": 1.901388168334961e-05, "eta_seconds": 39049.3523250178}
{"step": 1530, "total_steps": 57221, "tokens": 802160640, "source": "stack_v3", "source_tokens": {"stack_v3": 601882624, "fineweb_edu": 200278016}, "loss": 0.7945137545466423, "ce_loss": 0.7860568538308144, "perplexity": 2.1947252192322084, "z_loss": 84.56914138793945, "grad_norm": 0.10553229600191116, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165468.0274807988, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 25052.45847293828}
{"step": 1540, "total_steps": 57221, "tokens": 807403520, "source": "stack_v3", "source_tokens": {"stack_v3": 605552640, "fineweb_edu": 201850880}, "loss": 0.6888480186462402, "ce_loss": 0.6805652845650911, "perplexity": 1.9749938502782862, "z_loss": 82.82747077941895, "grad_norm": 0.1506214290857315, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1089624.2142126027, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 26791.43515647319}
{"step": 1550, "total_steps": 57221, "tokens": 812646400, "source": "stack_v3", "source_tokens": {"stack_v3": 609746944, "fineweb_edu": 202899456}, "loss": 0.7744732704013586, "ce_loss": 0.7661939095705748, "perplexity": 2.151561611825797, "z_loss": 82.79346466064453, "grad_norm": 0.10836713761091232, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1171820.7613599733, "data_wait_seconds": 1.704692840576172e-05, "eta_seconds": 24907.694557422074}
{"step": 1560, "total_steps": 57221, "tokens": 817889280, "source": "stack_v3", "source_tokens": {"stack_v3": 613416960, "fineweb_edu": 204472320}, "loss": 0.5023556090891361, "ce_loss": 0.49364236928522587, "perplexity": 1.6382725594275376, "z_loss": 87.13241577148438, "grad_norm": 0.09824319183826447, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1153414.795806566, "data_wait_seconds": 1.8596649169921875e-05, "eta_seconds": 25300.621100142365}
{"step": 1570, "total_steps": 57221, "tokens": 823132160, "source": "stack_v3", "source_tokens": {"stack_v3": 617611264, "fineweb_edu": 205520896}, "loss": 0.5204695761203766, "ce_loss": 0.5117021631449461, "perplexity": 1.6681282060864424, "z_loss": 87.6742172241211, "grad_norm": 0.13668197393417358, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150479.1182411113, "data_wait_seconds": 3.159046173095703e-05, "eta_seconds": 25360.62356751552}
{"step": 1580, "total_steps": 57221, "tokens": 828375040, "source": "stack_v3", "source_tokens": {"stack_v3": 621281280, "fineweb_edu": 207093760}, "loss": 0.34208450466394424, "ce_loss": 0.33273565769195557, "perplexity": 1.3947785507528228, "z_loss": 93.48851203918457, "grad_norm": 0.10445868223905563, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149844.7598154475, "data_wait_seconds": 2.1755695343017578e-05, "eta_seconds": 25370.055140906246}
{"step": 1590, "total_steps": 57221, "tokens": 833617920, "source": "stack_v3", "source_tokens": {"stack_v3": 625475584, "fineweb_edu": 208142336}, "loss": 0.5195413194596767, "ce_loss": 0.5110228266566992, "perplexity": 1.6669953705607947, "z_loss": 85.18502807617188, "grad_norm": 0.11190012842416763, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142175.4797992639, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 25535.81528919353}
{"step": 1600, "total_steps": 57221, "tokens": 838860800, "source": "stack_v3", "source_tokens": {"stack_v3": 629145600, "fineweb_edu": 209715200}, "loss": 0.6674472093582153, "ce_loss": 0.6595473885536194, "perplexity": 1.93391682339321, "z_loss": 78.99813842773438, "grad_norm": 0.1945498138666153, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 721682.4374715876, "data_wait_seconds": 0.7010036110877991, "eta_seconds": 40407.16205061879}
{"step": 1610, "total_steps": 57221, "tokens": 844103680, "source": "stack_v3", "source_tokens": {"stack_v3": 633339904, "fineweb_edu": 210763776}, "loss": 0.6014459654688835, "ce_loss": 0.5935934633016586, "perplexity": 1.8104826429611, "z_loss": 78.52494812011719, "grad_norm": 0.12825720012187958, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177921.8825638765, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 24751.97782771383}
{"step": 1620, "total_steps": 57221, "tokens": 849346560, "source": "stack_v3", "source_tokens": {"stack_v3": 637009920, "fineweb_edu": 212336640}, "loss": 0.4965241476893425, "ce_loss": 0.4885478764772415, "perplexity": 1.6299476153452526, "z_loss": 79.76255989074707, "grad_norm": 0.12939228117465973, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 438472.6588672929, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 66482.26029715268}
{"step": 1630, "total_steps": 57221, "tokens": 854589440, "source": "stack_v3", "source_tokens": {"stack_v3": 641204224, "fineweb_edu": 213385216}, "loss": 0.6159520354121923, "ce_loss": 0.6084076622501016, "perplexity": 1.837503142394578, "z_loss": 75.44380187988281, "grad_norm": 0.0968208834528923, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150454.3203063242, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 25333.826859148685}
{"step": 1640, "total_steps": 57221, "tokens": 859832320, "source": "stack_v3", "source_tokens": {"stack_v3": 644874240, "fineweb_edu": 214958080}, "loss": 0.6492391098290682, "ce_loss": 0.64216548204422, "perplexity": 1.900592124403437, "z_loss": 70.73639678955078, "grad_norm": 0.1326049566268921, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140513.6642630901, "data_wait_seconds": 1.7642974853515625e-05, "eta_seconds": 25550.038191631906}
{"step": 1650, "total_steps": 57221, "tokens": 865075200, "source": "stack_v3", "source_tokens": {"stack_v3": 649068544, "fineweb_edu": 216006656}, "loss": 0.5864562671631575, "ce_loss": 0.5795287657529116, "perplexity": 1.785196986549121, "z_loss": 69.27512073516846, "grad_norm": 0.11414165794849396, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1162125.926798573, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 25070.368131499272}
{"step": 1660, "total_steps": 57221, "tokens": 870318080, "source": "stack_v3", "source_tokens": {"stack_v3": 652738560, "fineweb_edu": 217579520}, "loss": 0.8742234613746405, "ce_loss": 0.8675324134528637, "perplexity": 2.3810282053695757, "z_loss": 66.91065120697021, "grad_norm": 0.19568878412246704, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157936.6532425017, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 25156.54188718344}
{"step": 1670, "total_steps": 57221, "tokens": 875560960, "source": "stack_v3", "source_tokens": {"stack_v3": 656932864, "fineweb_edu": 218628096}, "loss": 1.0077877910807729, "ce_loss": 1.0009186742827296, "perplexity": 2.720780191483281, "z_loss": 68.69144535064697, "grad_norm": 0.127625972032547, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1103400.8398863894, "data_wait_seconds": 1.8835067749023438e-05, "eta_seconds": 26395.157577547947}
{"step": 1680, "total_steps": 57221, "tokens": 880803840, "source": "stack_v3", "source_tokens": {"stack_v3": 660602880, "fineweb_edu": 220200960}, "loss": 0.8314208835363388, "ce_loss": 0.8242670930922031, "perplexity": 2.280208971862949, "z_loss": 71.53806114196777, "grad_norm": 0.19470398128032684, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1187086.7571761713, "data_wait_seconds": 1.6689300537109375e-05, "eta_seconds": 24529.964624715736}
{"step": 1690, "total_steps": 57221, "tokens": 886046720, "source": "stack_v3", "source_tokens": {"stack_v3": 664797184, "fineweb_edu": 221249536}, "loss": 0.9168720096349716, "ce_loss": 0.9102209396660328, "perplexity": 2.484871479415404, "z_loss": 66.51059246063232, "grad_norm": 0.17692819237709045, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 757775.5183727606, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 38420.287504825974}
{"step": 1700, "total_steps": 57221, "tokens": 891289600, "source": "stack_v3", "source_tokens": {"stack_v3": 668467200, "fineweb_edu": 222822400}, "loss": 0.9402194861322641, "ce_loss": 0.9341488070785999, "perplexity": 2.545046210178229, "z_loss": 60.70699501037598, "grad_norm": 0.518413245677948, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137806.1433493295, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 25583.18969373242}
{"step": 1710, "total_steps": 57221, "tokens": 896532480, "source": "stack_v3", "source_tokens": {"stack_v3": 672661504, "fineweb_edu": 223870976}, "loss": 0.7125980705022812, "ce_loss": 0.705476239323616, "perplexity": 2.0248107498432693, "z_loss": 71.21835422515869, "grad_norm": 0.14591214060783386, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149463.8795914035, "data_wait_seconds": 1.8835067749023438e-05, "eta_seconds": 25319.16664518882}
{"step": 1720, "total_steps": 57221, "tokens": 901775360, "source": "stack_v3", "source_tokens": {"stack_v3": 676331520, "fineweb_edu": 225443840}, "loss": 0.658842483535409, "ce_loss": 0.6525801680982113, "perplexity": 1.920489627972882, "z_loss": 62.623069763183594, "grad_norm": 0.21493230760097504, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149953.5947394075, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 25303.825104867807}
{"step": 1730, "total_steps": 57221, "tokens": 907018240, "source": "stack_v3", "source_tokens": {"stack_v3": 680525824, "fineweb_edu": 226492416}, "loss": 0.7262880988419056, "ce_loss": 0.7195558417588472, "perplexity": 2.0535209198182245, "z_loss": 67.32278251647949, "grad_norm": 0.11789603531360626, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148502.3147910754, "data_wait_seconds": 1.704692840576172e-05, "eta_seconds": 25331.23476141388}
{"step": 1740, "total_steps": 57221, "tokens": 912261120, "source": "stack_v3", "source_tokens": {"stack_v3": 684195840, "fineweb_edu": 228065280}, "loss": 0.7957228906452656, "ce_loss": 0.7895197812467813, "perplexity": 2.202338567993291, "z_loss": 62.03117561340332, "grad_norm": 0.1496317982673645, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138824.8878468731, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 25541.889003668446}
{"step": 1750, "total_steps": 57221, "tokens": 917504000, "source": "stack_v3", "source_tokens": {"stack_v3": 688390144, "fineweb_edu": 229113856}, "loss": 0.5664271004498005, "ce_loss": 0.5602902509272099, "perplexity": 1.7511807083629163, "z_loss": 61.36847400665283, "grad_norm": 0.10573720932006836, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166891.13060791, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24923.05857603787}
{"step": 1760, "total_steps": 57221, "tokens": 922746880, "source": "stack_v3", "source_tokens": {"stack_v3": 692060160, "fineweb_edu": 230686720}, "loss": 0.6681794142350554, "ce_loss": 0.6611952260136604, "perplexity": 1.9371062310704223, "z_loss": 69.84202575683594, "grad_norm": 0.12351097911596298, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152806.8692261644, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 25223.004734104732}
{"step": 1770, "total_steps": 57221, "tokens": 927989760, "source": "stack_v3", "source_tokens": {"stack_v3": 696254464, "fineweb_edu": 231735296}, "loss": 0.555677792057395, "ce_loss": 0.5493138916790485, "perplexity": 1.7320642264160102, "z_loss": 63.63900566101074, "grad_norm": 0.104483962059021, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 451445.4241823802, "data_wait_seconds": 2.1457672119140625e-05, "eta_seconds": 64397.618588454556}
{"step": 1780, "total_steps": 57221, "tokens": 933232640, "source": "stack_v3", "source_tokens": {"stack_v3": 699924480, "fineweb_edu": 233308160}, "loss": 0.4046306721866131, "ce_loss": 0.3989196829497814, "perplexity": 1.490213924214129, "z_loss": 57.10996627807617, "grad_norm": 0.08451581746339798, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147712.991045689, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 25325.815414459215}
{"step": 1790, "total_steps": 57221, "tokens": 938475520, "source": "stack_v3", "source_tokens": {"stack_v3": 704118784, "fineweb_edu": 234356736}, "loss": 0.490145742893219, "ce_loss": 0.4846283784136176, "perplexity": 1.6235715424883872, "z_loss": 55.173584938049316, "grad_norm": 0.15588493645191193, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1041327.8013408126, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 27908.142318471102}
{"step": 1800, "total_steps": 57221, "tokens": 943718400, "source": "stack_v3", "source_tokens": {"stack_v3": 707788800, "fineweb_edu": 235929600}, "loss": 0.7607500664889812, "ce_loss": 0.7548697777092457, "perplexity": 2.1273344789479616, "z_loss": 58.80270862579346, "grad_norm": 0.17043589055538177, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1121778.3102565212, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 25901.981999772837}
{"step": 1810, "total_steps": 57221, "tokens": 948961280, "source": "stack_v3", "source_tokens": {"stack_v3": 711983104, "fineweb_edu": 236978176}, "loss": 0.4225853653624654, "ce_loss": 0.41627524327486753, "perplexity": 1.5163031636876039, "z_loss": 63.10109043121338, "grad_norm": 0.09010192006826401, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150693.7974712425, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 25246.541507256214}
{"step": 1820, "total_steps": 57221, "tokens": 954204160, "source": "stack_v3", "source_tokens": {"stack_v3": 715653120, "fineweb_edu": 238551040}, "loss": 0.5168777741491795, "ce_loss": 0.5107725281268358, "perplexity": 1.6665781762839893, "z_loss": 61.052534103393555, "grad_norm": 0.09507706016302109, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1176639.2101971752, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 24685.38834018004}
{"step": 1830, "total_steps": 57221, "tokens": 959447040, "source": "stack_v3", "source_tokens": {"stack_v3": 719847424, "fineweb_edu": 239599616}, "loss": 0.7050855765119195, "ce_loss": 0.6991310669109225, "perplexity": 2.0120036511262533, "z_loss": 59.54509449005127, "grad_norm": 0.11014694720506668, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135889.922403371, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 25566.344402945833}
{"step": 1840, "total_steps": 57221, "tokens": 964689920, "source": "stack_v3", "source_tokens": {"stack_v3": 723517440, "fineweb_edu": 241172480}, "loss": 0.6497890967875719, "ce_loss": 0.644747487269342, "perplexity": 1.9055058040428678, "z_loss": 50.416022300720215, "grad_norm": 0.10759517550468445, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148540.405763178, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 25280.181641242933}
{"step": 1850, "total_steps": 57221, "tokens": 969932800, "source": "stack_v3", "source_tokens": {"stack_v3": 727711744, "fineweb_edu": 242221056}, "loss": 0.8848516056314111, "ce_loss": 0.8799171661958098, "perplexity": 2.410700010693899, "z_loss": 49.344242095947266, "grad_norm": 0.1662597954273224, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148257.3932201106, "data_wait_seconds": 2.1576881408691406e-05, "eta_seconds": 25281.84653668082}
{"step": 1860, "total_steps": 57221, "tokens": 975175680, "source": "stack_v3", "source_tokens": {"stack_v3": 731381760, "fineweb_edu": 243793920}, "loss": 0.7684744689613581, "ce_loss": 0.7636269051581621, "perplexity": 2.14604562648299, "z_loss": 48.47568130493164, "grad_norm": 0.13640838861465454, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1085826.0392511839, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 26730.63941256772}
{"step": 1870, "total_steps": 57221, "tokens": 980418560, "source": "stack_v3", "source_tokens": {"stack_v3": 735576064, "fineweb_edu": 244842496}, "loss": 0.7574698403477669, "ce_loss": 0.7522232457995415, "perplexity": 2.121711863863724, "z_loss": 52.465993881225586, "grad_norm": 0.088487409055233, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1162095.281373374, "data_wait_seconds": 1.4901161193847656e-05, "eta_seconds": 24971.77460845071}
{"step": 1880, "total_steps": 57221, "tokens": 985661440, "source": "stack_v3", "source_tokens": {"stack_v3": 739246080, "fineweb_edu": 246415360}, "loss": 0.6760666072368622, "ce_loss": 0.670871764421463, "perplexity": 1.9559416980029998, "z_loss": 51.94869613647461, "grad_norm": 0.1340664029121399, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155348.9492856036, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 25113.05227562692}
{"step": 1890, "total_steps": 57221, "tokens": 990904320, "source": "stack_v3", "source_tokens": {"stack_v3": 743440384, "fineweb_edu": 247463936}, "loss": 0.7339601218700409, "ce_loss": 0.7285628914833069, "perplexity": 2.0721006334460546, "z_loss": 53.972307205200195, "grad_norm": 0.18269279599189758, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154619.7219837427, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 25124.372230676705}
{"step": 1900, "total_steps": 57221, "tokens": 996147200, "source": "stack_v3", "source_tokens": {"stack_v3": 747110400, "fineweb_edu": 249036800}, "loss": 1.478139966726303, "ce_loss": 1.472727656364441, "perplexity": 4.3611145540886405, "z_loss": 54.12341499328613, "grad_norm": 0.19486035406589508, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1179971.7794687725, "data_wait_seconds": 1.6689300537109375e-05, "eta_seconds": 24580.124122169804}
{"step": 1910, "total_steps": 57221, "tokens": 1001390080, "source": "stack_v3", "source_tokens": {"stack_v3": 751304704, "fineweb_edu": 250085376}, "loss": 0.9442152082920074, "ce_loss": 0.9394469261169434, "perplexity": 2.55856595093079, "z_loss": 47.68276786804199, "grad_norm": 0.12894576787948608, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132045.949126321, "data_wait_seconds": 1.710653305053711e-05, "eta_seconds": 25616.10678646062}
{"step": 1920, "total_steps": 57221, "tokens": 1006632960, "source": "stack_v3", "source_tokens": {"stack_v3": 754974720, "fineweb_edu": 251658240}, "loss": 0.9622541069984436, "ce_loss": 0.9559376239776611, "perplexity": 2.6011083014211778, "z_loss": 63.16469955444336, "grad_norm": 0.9152310490608215, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137587.4180913388, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 25486.71563952905}
{"step": 1930, "total_steps": 57221, "tokens": 1011875840, "source": "stack_v3", "source_tokens": {"stack_v3": 759169024, "fineweb_edu": 252706816}, "loss": 0.8648044839501381, "ce_loss": 0.8597893565893173, "perplexity": 2.3626629619010844, "z_loss": 50.15121555328369, "grad_norm": 0.16705146431922913, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1045533.5036465151, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 27725.676947604166}
{"step": 1940, "total_steps": 57221, "tokens": 1017118720, "source": "stack_v3", "source_tokens": {"stack_v3": 762839040, "fineweb_edu": 254279680}, "loss": 0.8166590183973312, "ce_loss": 0.8107803240418434, "perplexity": 2.249662767882267, "z_loss": 58.78692626953125, "grad_norm": 0.19925624132156372, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1179880.0417283585, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 24564.26098838332}
{"step": 1950, "total_steps": 57221, "tokens": 1022361600, "source": "stack_v3", "source_tokens": {"stack_v3": 767033344, "fineweb_edu": 255328256}, "loss": 0.9698052182793617, "ce_loss": 0.9651363044977188, "perplexity": 2.6251454512216723, "z_loss": 46.68930912017822, "grad_norm": 0.2124318927526474, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1068829.2280981394, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 27111.569966665702}
{"step": 1960, "total_steps": 57221, "tokens": 1027604480, "source": "stack_v3", "source_tokens": {"stack_v3": 770703360, "fineweb_edu": 256901120}, "loss": 0.5921431183815002, "ce_loss": 0.5874202698469162, "perplexity": 1.799340609706683, "z_loss": 47.22847843170166, "grad_norm": 0.12724794447422028, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152230.795597781, "data_wait_seconds": 2.300739288330078e-05, "eta_seconds": 25144.611331941556}
{"step": 1970, "total_steps": 57221, "tokens": 1032847360, "source": "stack_v3", "source_tokens": {"stack_v3": 774897664, "fineweb_edu": 257949696}, "loss": 0.9675219506025314, "ce_loss": 0.9626658856868744, "perplexity": 2.618668246516126, "z_loss": 48.56090450286865, "grad_norm": 0.14462564885616302, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138609.3080569918, "data_wait_seconds": 2.4020671844482422e-05, "eta_seconds": 25440.818404542748}
{"step": 1980, "total_steps": 57221, "tokens": 1038090240, "source": "stack_v3", "source_tokens": {"stack_v3": 778567680, "fineweb_edu": 259522560}, "loss": 0.911147803068161, "ce_loss": 0.905945710837841, "perplexity": 2.474270761625688, "z_loss": 52.02089309692383, "grad_norm": 0.15973997116088867, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1127810.5369890763, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 25679.765182297208}
{"step": 1990, "total_steps": 57221, "tokens": 1043333120, "source": "stack_v3", "source_tokens": {"stack_v3": 782761984, "fineweb_edu": 260571136}, "loss": 1.1392710208892822, "ce_loss": 1.133901208639145, "perplexity": 3.1077568897323125, "z_loss": 53.69776725769043, "grad_norm": 0.12144055217504501, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145878.8714895586, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 25270.26861256151}
{"step": 2000, "total_steps": 57221, "tokens": 1048576000, "source": "stack_v3", "source_tokens": {"stack_v3": 786432000, "fineweb_edu": 262144000}, "loss": 1.2061851024627686, "ce_loss": 1.2014914453029633, "perplexity": 3.3250723900118655, "z_loss": 46.93633270263672, "grad_norm": 0.2094135880470276, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126532.7550929347, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 25699.58473831648}
{"step": 2010, "total_steps": 57221, "tokens": 1053818880, "source": "stack_v3", "source_tokens": {"stack_v3": 790626304, "fineweb_edu": 263192576}, "loss": 1.1433230489492416, "ce_loss": 1.138571947813034, "perplexity": 3.122306363528021, "z_loss": 47.51073741912842, "grad_norm": 0.3014988303184509, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1124716.122428991, "data_wait_seconds": 1.627206802368164e-05, "eta_seconds": 25736.43299207487}
{"step": 2020, "total_steps": 57221, "tokens": 1059061760, "source": "stack_v3", "source_tokens": {"stack_v3": 794296320, "fineweb_edu": 264765440}, "loss": 0.7976932674646378, "ce_loss": 0.7925202921032906, "perplexity": 2.208956632598374, "z_loss": 51.729796409606934, "grad_norm": 0.13917872309684753, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132690.4531646208, "data_wait_seconds": 2.187490463256836e-05, "eta_seconds": 25550.61549176299}
{"step": 2030, "total_steps": 57221, "tokens": 1064304640, "source": "stack_v3", "source_tokens": {"stack_v3": 798490624, "fineweb_edu": 265814016}, "loss": 1.4609062746167183, "ce_loss": 1.4560093730688095, "perplexity": 4.288810291149286, "z_loss": 48.96906852722168, "grad_norm": 0.18909215927124023, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1052961.5153828198, "data_wait_seconds": 1.7583370208740234e-05, "eta_seconds": 27480.297178268665}
{"step": 2040, "total_steps": 57221, "tokens": 1069547520, "source": "stack_v3", "source_tokens": {"stack_v3": 802160640, "fineweb_edu": 267386880}, "loss": 0.763721888884902, "ce_loss": 0.7590266838669777, "perplexity": 2.136196014267915, "z_loss": 46.952120780944824, "grad_norm": 0.11553916335105896, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140778.2544295604, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 25360.276958002243}
{"step": 2050, "total_steps": 57221, "tokens": 1074790400, "source": "stack_v3", "source_tokens": {"stack_v3": 806354944, "fineweb_edu": 268435456}, "loss": 1.0664749294519424, "ce_loss": 1.0613868534564972, "perplexity": 2.890376739903673, "z_loss": 50.88077735900879, "grad_norm": 0.11099810153245926, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1156191.2442873241, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 25017.66878353199}
{"step": 2060, "total_steps": 57221, "tokens": 1080033280, "source": "stack_v3", "source_tokens": {"stack_v3": 810024960, "fineweb_edu": 270008320}, "loss": 1.0524652302265167, "ce_loss": 1.047780603170395, "perplexity": 2.851315889025702, "z_loss": 46.846120834350586, "grad_norm": 0.2397005409002304, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108822.6331914351, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 26081.6886797864}
{"step": 2070, "total_steps": 57221, "tokens": 1085276160, "source": "stack_v3", "source_tokens": {"stack_v3": 814219264, "fineweb_edu": 271056896}, "loss": 0.7560401856899261, "ce_loss": 0.7506479322910309, "perplexity": 2.1183721337550496, "z_loss": 53.922555923461914, "grad_norm": 0.14759549498558044, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1103332.6341880886, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 26206.7149507252}
{"step": 2080, "total_steps": 57221, "tokens": 1090519040, "source": "stack_v3", "source_tokens": {"stack_v3": 817889280, "fineweb_edu": 272629760}, "loss": 0.8348444402217865, "ce_loss": 0.8302973061800003, "perplexity": 2.294000659462747, "z_loss": 45.4712610244751, "grad_norm": 0.1459059864282608, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1173635.011564906, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 24632.428885580506}
{"step": 2090, "total_steps": 57221, "tokens": 1095761920, "source": "stack_v3", "source_tokens": {"stack_v3": 822083584, "fineweb_edu": 273678336}, "loss": 1.0534173101186752, "ce_loss": 1.0490538477897644, "perplexity": 2.8549486238287796, "z_loss": 43.63477897644043, "grad_norm": 0.10251954942941666, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126771.0360660274, "data_wait_seconds": 1.710653305053711e-05, "eta_seconds": 25652.27287072921}
{"step": 2100, "total_steps": 57221, "tokens": 1101004800, "source": "stack_v3", "source_tokens": {"stack_v3": 825753600, "fineweb_edu": 275251200}, "loss": 1.1423668712377548, "ce_loss": 1.138207957148552, "perplexity": 3.121170079970893, "z_loss": 41.58905220031738, "grad_norm": 0.20471514761447906, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134460.8497274693, "data_wait_seconds": 1.7464160919189453e-05, "eta_seconds": 25473.77038788283}
{"step": 2110, "total_steps": 57221, "tokens": 1106247680, "source": "stack_v3", "source_tokens": {"stack_v3": 829947904, "fineweb_edu": 276299776}, "loss": 0.8197289109230042, "ce_loss": 0.8154283761978149, "perplexity": 2.260143656750355, "z_loss": 43.00548839569092, "grad_norm": 0.21969164907932281, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1077628.3369013774, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 26812.353879892733}
{"step": 2120, "total_steps": 57221, "tokens": 1111490560, "source": "stack_v3", "source_tokens": {"stack_v3": 833617920, "fineweb_edu": 277872640}, "loss": 1.055277019739151, "ce_loss": 1.051195189356804, "perplexity": 2.861068594123623, "z_loss": 40.81831073760986, "grad_norm": 0.39472708106040955, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177192.9419448199, "data_wait_seconds": 2.1159648895263672e-05, "eta_seconds": 24540.165346450172}
{"step": 2130, "total_steps": 57221, "tokens": 1116733440, "source": "stack_v3", "source_tokens": {"stack_v3": 837812224, "fineweb_edu": 278921216}, "loss": 1.0030692666769028, "ce_loss": 0.9983942061662674, "perplexity": 2.7139203310306, "z_loss": 46.75050163269043, "grad_norm": 0.10096346586942673, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141578.8660696084, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 25301.157386912266}
{"step": 2140, "total_steps": 57221, "tokens": 1121976320, "source": "stack_v3", "source_tokens": {"stack_v3": 841482240, "fineweb_edu": 280494080}, "loss": 0.7619640529155731, "ce_loss": 0.7579703480005264, "perplexity": 2.133940665213041, "z_loss": 39.93727207183838, "grad_norm": 0.19657601416110992, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126860.5329097188, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 25626.972315227613}
{"step": 2150, "total_steps": 57221, "tokens": 1127219200, "source": "stack_v3", "source_tokens": {"stack_v3": 845676544, "fineweb_edu": 281542656}, "loss": 1.345434531569481, "ce_loss": 1.3412901014089584, "perplexity": 3.823973638275395, "z_loss": 41.44429683685303, "grad_norm": 0.12938083708286285, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149269.720011799, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 25122.719495040365}
{"step": 2160, "total_steps": 57221, "tokens": 1132462080, "source": "stack_v3", "source_tokens": {"stack_v3": 849346560, "fineweb_edu": 283115520}, "loss": 1.060264527797699, "ce_loss": 1.056131511926651, "perplexity": 2.8752266672571873, "z_loss": 41.330397605895996, "grad_norm": 0.16805601119995117, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1106185.5772190357, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 26096.469267456323}
{"step": 2170, "total_steps": 57221, "tokens": 1137704960, "source": "stack_v3", "source_tokens": {"stack_v3": 853540864, "fineweb_edu": 284164096}, "loss": 0.9326661676168442, "ce_loss": 0.9284974038600922, "perplexity": 2.5307036936797433, "z_loss": 41.68748378753662, "grad_norm": 0.17090077698230743, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1081753.454885258, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 26681.028759054374}
{"step": 2180, "total_steps": 57221, "tokens": 1142947840, "source": "stack_v3", "source_tokens": {"stack_v3": 857210880, "fineweb_edu": 285736960}, "loss": 0.8439444899559021, "ce_loss": 0.8400118201971054, "perplexity": 2.3163943568571446, "z_loss": 39.326629638671875, "grad_norm": 0.1630886048078537, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 848482.5829095499, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 34010.18800061359}
{"step": 2190, "total_steps": 57221, "tokens": 1148190720, "source": "stack_v3", "source_tokens": {"stack_v3": 861405184, "fineweb_edu": 286785536}, "loss": 0.8499074876308441, "ce_loss": 0.8455568253993988, "perplexity": 2.329274452713648, "z_loss": 43.50664043426514, "grad_norm": 0.14551439881324768, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1117016.5618695419, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 25829.347804575926}
{"step": 2200, "total_steps": 57221, "tokens": 1153433600, "source": "stack_v3", "source_tokens": {"stack_v3": 865075200, "fineweb_edu": 288358400}, "loss": 0.743940994143486, "ce_loss": 0.7402553558349609, "perplexity": 2.0964707921979264, "z_loss": 36.85624361038208, "grad_norm": 0.1296924203634262, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140888.1025005782, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 25284.308195321355}
{"step": 2210, "total_steps": 57221, "tokens": 1158676480, "source": "stack_v3", "source_tokens": {"stack_v3": 869269504, "fineweb_edu": 289406976}, "loss": 0.8495504073798656, "ce_loss": 0.8453928660601377, "perplexity": 2.3288925777202647, "z_loss": 41.575300216674805, "grad_norm": 0.2436760514974594, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143618.5359268463, "data_wait_seconds": 2.4437904357910156e-05, "eta_seconds": 25219.356467168076}
{"step": 2220, "total_steps": 57221, "tokens": 1163919360, "source": "stack_v3", "source_tokens": {"stack_v3": 872939520, "fineweb_edu": 290979840}, "loss": 0.9040846228599548, "ce_loss": 0.9000874049961567, "perplexity": 2.459818102152934, "z_loss": 39.972018241882324, "grad_norm": 0.1369396150112152, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1159591.3301410722, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 24867.451049752068}
{"step": 2230, "total_steps": 57221, "tokens": 1169162240, "source": "stack_v3", "source_tokens": {"stack_v3": 877133824, "fineweb_edu": 292028416}, "loss": 0.8461862839758396, "ce_loss": 0.8420540913939476, "perplexity": 2.3211298963146554, "z_loss": 41.32209348678589, "grad_norm": 0.12422288954257965, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129211.7602705192, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 25531.825627722075}
{"step": 2240, "total_steps": 57221, "tokens": 1174405120, "source": "stack_v3", "source_tokens": {"stack_v3": 880803840, "fineweb_edu": 293601280}, "loss": 0.6172478422522545, "ce_loss": 0.6135610342025757, "perplexity": 1.8469969210267345, "z_loss": 36.867862701416016, "grad_norm": 0.09320934861898422, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119257.740726559, "data_wait_seconds": 1.8596649169921875e-05, "eta_seconds": 25754.20640940848}
{"step": 2250, "total_steps": 57221, "tokens": 1179648000, "source": "stack_v3", "source_tokens": {"stack_v3": 884998144, "fineweb_edu": 294649856}, "loss": 0.6832897290587425, "ce_loss": 0.6789976507425308, "perplexity": 1.9719002086014508, "z_loss": 42.920857429504395, "grad_norm": 0.09422086179256439, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141334.6963905348, "data_wait_seconds": 2.09808349609375e-05, "eta_seconds": 25251.44647853449}
{"step": 2260, "total_steps": 57221, "tokens": 1184890880, "source": "stack_v3", "source_tokens": {"stack_v3": 888668160, "fineweb_edu": 296222720}, "loss": 0.6421228386461735, "ce_loss": 0.6380590386688709, "perplexity": 1.8928034532783593, "z_loss": 40.6379280090332, "grad_norm": 0.18141892552375793, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154944.8214019358, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 24949.33834589832}
{"step": 2270, "total_steps": 57221, "tokens": 1190133760, "source": "stack_v3", "source_tokens": {"stack_v3": 892862464, "fineweb_edu": 297271296}, "loss": 0.7943493276834488, "ce_loss": 0.7902875766158104, "perplexity": 2.204030162663006, "z_loss": 40.617465019226074, "grad_norm": 0.12740233540534973, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134333.6868792407, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 25398.052242688136}
{"step": 2280, "total_steps": 57221, "tokens": 1195376640, "source": "stack_v3", "source_tokens": {"stack_v3": 896532480, "fineweb_edu": 298844160}, "loss": 0.8415623009204865, "ce_loss": 0.8383707255125046, "perplexity": 2.312596051932142, "z_loss": 31.91582202911377, "grad_norm": 0.24104583263397217, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128047.275685375, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 25534.943420256022}
{"step": 2290, "total_steps": 57221, "tokens": 1200619520, "source": "stack_v3", "source_tokens": {"stack_v3": 900726784, "fineweb_edu": 299892736}, "loss": 0.762919694185257, "ce_loss": 0.7596691027283669, "perplexity": 2.137568787779691, "z_loss": 32.505910873413086, "grad_norm": 0.15430468320846558, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099469.518093457, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 26193.887148358393}
{"step": 2300, "total_steps": 57221, "tokens": 1205862400, "source": "stack_v3", "source_tokens": {"stack_v3": 904396800, "fineweb_edu": 301465600}, "loss": 0.8074838109314442, "ce_loss": 0.8046119622886181, "perplexity": 2.235828744601098, "z_loss": 28.718698024749756, "grad_norm": 0.11248379945755005, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166181.402213568, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 24690.95935276011}
{"step": 2310, "total_steps": 57221, "tokens": 1211105280, "source": "stack_v3", "source_tokens": {"stack_v3": 908591104, "fineweb_edu": 302514176}, "loss": 0.9636463522911072, "ce_loss": 0.9601094424724579, "perplexity": 2.611982319584049, "z_loss": 35.36890697479248, "grad_norm": 0.14033302664756775, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1103178.0399919131, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 26096.326863260474}
{"step": 2320, "total_steps": 57221, "tokens": 1216348160, "source": "stack_v3", "source_tokens": {"stack_v3": 912261120, "fineweb_edu": 304087040}, "loss": 1.0902107059955597, "ce_loss": 1.0867213010787964, "perplexity": 2.964538292437413, "z_loss": 34.8939323425293, "grad_norm": 0.18536072969436646, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152905.869063469, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 24966.17686869926}
{"step": 2330, "total_steps": 57221, "tokens": 1221591040, "source": "stack_v3", "source_tokens": {"stack_v3": 916455424, "fineweb_edu": 305135616}, "loss": 0.7724679261445999, "ce_loss": 0.7691064924001694, "perplexity": 2.1578373480975754, "z_loss": 33.61435127258301, "grad_norm": 0.10101857036352158, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1167491.1960212917, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 24649.786703380985}
{"step": 2340, "total_steps": 57221, "tokens": 1226833920, "source": "stack_v3", "source_tokens": {"stack_v3": 920125440, "fineweb_edu": 306708480}, "loss": 0.8244999349117279, "ce_loss": 0.8205033391714096, "perplexity": 2.2716429567035092, "z_loss": 39.96590518951416, "grad_norm": 0.1237613782286644, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149970.9140195346, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 25020.77724681585}
{"step": 2350, "total_steps": 57221, "tokens": 1232076800, "source": "stack_v3", "source_tokens": {"stack_v3": 924319744, "fineweb_edu": 307757056}, "loss": 1.086046352982521, "ce_loss": 1.0826451033353806, "perplexity": 2.952478843187118, "z_loss": 34.01244497299194, "grad_norm": 0.15188869833946228, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1092464.5215103196, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 26333.050303756023}
{"step": 2360, "total_steps": 57221, "tokens": 1237319680, "source": "stack_v3", "source_tokens": {"stack_v3": 927989760, "fineweb_edu": 309329920}, "loss": 1.0170102566480637, "ce_loss": 1.0132098495960236, "perplexity": 2.754428140619665, "z_loss": 38.00410842895508, "grad_norm": 0.15464699268341064, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135750.2377623303, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 25324.82879041135}
{"step": 2370, "total_steps": 57221, "tokens": 1242562560, "source": "stack_v3", "source_tokens": {"stack_v3": 932184064, "fineweb_edu": 310378496}, "loss": 0.9789162129163742, "ce_loss": 0.9756078124046326, "perplexity": 2.652779113117629, "z_loss": 33.083964824676514, "grad_norm": 0.11261036247015, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1175150.4823607716, "data_wait_seconds": 2.2292137145996094e-05, "eta_seconds": 24471.28080331371}
{"step": 2380, "total_steps": 57221, "tokens": 1247805440, "source": "stack_v3", "source_tokens": {"stack_v3": 935854080, "fineweb_edu": 311951360}, "loss": 1.19440758228302, "ce_loss": 1.1900694519281387, "perplexity": 3.287309509438837, "z_loss": 43.381330490112305, "grad_norm": 0.18807387351989746, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119881.317033724, "data_wait_seconds": 1.621246337890625e-05, "eta_seconds": 25674.32291500061}
{"step": 2390, "total_steps": 57221, "tokens": 1253048320, "source": "stack_v3", "source_tokens": {"stack_v3": 940048384, "fineweb_edu": 312999936}, "loss": 0.8631943166255951, "ce_loss": 0.8597995638847351, "perplexity": 2.362687078422991, "z_loss": 33.947561264038086, "grad_norm": 0.1374392807483673, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 723854.4673402556, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 39713.71729683224}
{"step": 2400, "total_steps": 57221, "tokens": 1258291200, "source": "stack_v3", "source_tokens": {"stack_v3": 943718400, "fineweb_edu": 314572800}, "loss": 0.8894054740667343, "ce_loss": 0.8863669484853745, "perplexity": 2.4262987510929133, "z_loss": 30.3851261138916, "grad_norm": 0.13137708604335785, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145531.9416154746, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 25090.272698522313}
{"step": 2410, "total_steps": 57221, "tokens": 1263534080, "source": "stack_v3", "source_tokens": {"stack_v3": 947912704, "fineweb_edu": 315621376}, "loss": 0.7504288405179977, "ce_loss": 0.7466209605336189, "perplexity": 2.1098586622602538, "z_loss": 38.07893466949463, "grad_norm": 0.12852172553539276, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168333.0452212363, "data_wait_seconds": 2.2590160369873047e-05, "eta_seconds": 24596.12525515654}
{"step": 2420, "total_steps": 57221, "tokens": 1268776960, "source": "stack_v3", "source_tokens": {"stack_v3": 951582720, "fineweb_edu": 317194240}, "loss": 0.8102118074893951, "ce_loss": 0.8067449554800987, "perplexity": 2.2406028418390576, "z_loss": 34.66862392425537, "grad_norm": 0.16982968151569366, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142495.270333655, "data_wait_seconds": 1.7404556274414062e-05, "eta_seconds": 25147.782915205695}
{"step": 2430, "total_steps": 57221, "tokens": 1274019840, "source": "stack_v3", "source_tokens": {"stack_v3": 955777024, "fineweb_edu": 318242816}, "loss": 0.8475907817482948, "ce_loss": 0.8441374078392982, "perplexity": 2.3259705849918078, "z_loss": 34.5338830947876, "grad_norm": 0.1287069469690323, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136261.4477140538, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 25281.13597252755}
{"step": 2440, "total_steps": 57221, "tokens": 1279262720, "source": "stack_v3", "source_tokens": {"stack_v3": 959447040, "fineweb_edu": 319815680}, "loss": 0.9041948392987251, "ce_loss": 0.900665894150734, "perplexity": 2.4612414919153376, "z_loss": 35.289451599121094, "grad_norm": 0.13789252936840057, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168877.9286830414, "data_wait_seconds": 2.6524066925048828e-05, "eta_seconds": 24571.203352568442}
{"step": 2450, "total_steps": 57221, "tokens": 1284505600, "source": "stack_v3", "source_tokens": {"stack_v3": 963641344, "fineweb_edu": 320864256}, "loss": 0.9548513516783714, "ce_loss": 0.9511895179748535, "perplexity": 2.5887872374877463, "z_loss": 36.61833477020264, "grad_norm": 0.16345201432704926, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1183659.2033869412, "data_wait_seconds": 1.6868114471435547e-05, "eta_seconds": 24259.934208962368}
{"step": 2460, "total_steps": 57221, "tokens": 1289748480, "source": "stack_v3", "source_tokens": {"stack_v3": 967311360, "fineweb_edu": 322437120}, "loss": 0.6710045859217644, "ce_loss": 0.6673626825213432, "perplexity": 1.949090166715677, "z_loss": 36.41904258728027, "grad_norm": 0.13905340433120728, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1078983.3756140855, "data_wait_seconds": 2.1159648895263672e-05, "eta_seconds": 26608.61341228732}
{"step": 2470, "total_steps": 57221, "tokens": 1294991360, "source": "stack_v3", "source_tokens": {"stack_v3": 971505664, "fineweb_edu": 323485696}, "loss": 0.7050744816660881, "ce_loss": 0.7021446079015732, "perplexity": 2.0180760517208074, "z_loss": 29.29875421524048, "grad_norm": 0.10535935312509537, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1084201.1356489896, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 26475.72271986003}
{"step": 2480, "total_steps": 57221, "tokens": 1300234240, "source": "stack_v3", "source_tokens": {"stack_v3": 975175680, "fineweb_edu": 325058560}, "loss": 0.8080054670572281, "ce_loss": 0.8048277199268341, "perplexity": 2.236311193774664, "z_loss": 31.777477264404297, "grad_norm": 0.1766793429851532, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1172519.5544771405, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 24477.003944550874}
{"step": 2490, "total_steps": 57221, "tokens": 1305477120, "source": "stack_v3", "source_tokens": {"stack_v3": 979369984, "fineweb_edu": 326107136}, "loss": 0.9520274251699448, "ce_loss": 0.9487616419792175, "perplexity": 2.582509606834095, "z_loss": 32.657716274261475, "grad_norm": 0.09451057016849518, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108903.7095245603, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 25876.47839351417}
{"step": 2500, "total_steps": 57221, "tokens": 1310720000, "source": "stack_v3", "source_tokens": {"stack_v3": 983040000, "fineweb_edu": 327680000}, "loss": 1.0721096247434616, "ce_loss": 1.06912861764431, "perplexity": 2.9128401962334913, "z_loss": 29.810133934020996, "grad_norm": 0.1854160875082016, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099670.9151080535, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 26088.96862310939}
{"step": 2510, "total_steps": 57221, "tokens": 1315962880, "source": "stack_v3", "source_tokens": {"stack_v3": 987234304, "fineweb_edu": 328728576}, "loss": 1.1236463338136673, "ce_loss": 1.1202906668186188, "perplexity": 3.065745184196823, "z_loss": 33.556546688079834, "grad_norm": 0.21466663479804993, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1070536.9041432212, "data_wait_seconds": 1.4841556549072266e-05, "eta_seconds": 26794.066611796618}
{"step": 2520, "total_steps": 57221, "tokens": 1321205760, "source": "stack_v3", "source_tokens": {"stack_v3": 990904320, "fineweb_edu": 330301440}, "loss": 1.4728345572948456, "ce_loss": 1.4694707989692688, "perplexity": 4.346934130245547, "z_loss": 33.63776397705078, "grad_norm": 0.20062102377414703, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108538.568190633, "data_wait_seconds": 1.7821788787841797e-05, "eta_seconds": 25870.813215646427}
{"step": 2530, "total_steps": 57221, "tokens": 1326448640, "source": "stack_v3", "source_tokens": {"stack_v3": 995098624, "fineweb_edu": 331350016}, "loss": 0.950468398630619, "ce_loss": 0.9469368606805801, "perplexity": 2.5778013886407094, "z_loss": 35.31535053253174, "grad_norm": 0.11285323649644852, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 884883.669027713, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 32403.752452009594}
{"step": 2540, "total_steps": 57221, "tokens": 1331691520, "source": "stack_v3", "source_tokens": {"stack_v3": 998768640, "fineweb_edu": 332922880}, "loss": 1.1354003548622131, "ce_loss": 1.1314856708049774, "perplexity": 3.100259044695666, "z_loss": 39.14696407318115, "grad_norm": 0.5397785902023315, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130478.1370986614, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 25359.45414528437}
{"step": 2550, "total_steps": 57221, "tokens": 1336934400, "source": "stack_v3", "source_tokens": {"stack_v3": 1002962944, "fineweb_edu": 333971456}, "loss": 1.06788469851017, "ce_loss": 1.0643327981233597, "perplexity": 2.898904184368405, "z_loss": 35.51912212371826, "grad_norm": 0.14355242252349854, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1078272.886726155, "data_wait_seconds": 2.1636486053466797e-05, "eta_seconds": 26582.385547156446}
{"step": 2560, "total_steps": 57221, "tokens": 1342177280, "source": "stack_v3", "source_tokens": {"stack_v3": 1006632960, "fineweb_edu": 335544320}, "loss": 0.8992584049701691, "ce_loss": 0.8958242684602737, "perplexity": 2.4493538827763377, "z_loss": 34.34139013290405, "grad_norm": 0.234770268201828, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151530.6427242989, "data_wait_seconds": 1.7702579498291016e-05, "eta_seconds": 24886.721774247475}
{"step": 2570, "total_steps": 57221, "tokens": 1347420160, "source": "stack_v3", "source_tokens": {"stack_v3": 1010827264, "fineweb_edu": 336592896}, "loss": 0.9548708349466324, "ce_loss": 0.9502538442611694, "perplexity": 2.5863661101885325, "z_loss": 46.16989517211914, "grad_norm": 0.33347445726394653, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129816.1787205406, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 25360.39081370528}
{"step": 2580, "total_steps": 57221, "tokens": 1352663040, "source": "stack_v3", "source_tokens": {"stack_v3": 1014497280, "fineweb_edu": 338165760}, "loss": 0.7743076086044312, "ce_loss": 0.7698192223906517, "perplexity": 2.1593758516938464, "z_loss": 44.883750915527344, "grad_norm": 0.11598502844572067, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143119.3445311752, "data_wait_seconds": 1.621246337890625e-05, "eta_seconds": 25060.670259017497}
{"step": 2590, "total_steps": 57221, "tokens": 1357905920, "source": "stack_v3", "source_tokens": {"stack_v3": 1018691584, "fineweb_edu": 339214336}, "loss": 0.6669460982084274, "ce_loss": 0.6637582369148731, "perplexity": 1.9420774233461369, "z_loss": 31.878662109375, "grad_norm": 0.9410489201545715, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177212.792988918, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 24330.43053098186}
{"step": 2600, "total_steps": 57221, "tokens": 1363148800, "source": "stack_v3", "source_tokens": {"stack_v3": 1022361600, "fineweb_edu": 340787200}, "loss": 0.7996299639344215, "ce_loss": 0.7963671833276749, "perplexity": 2.2174706141642857, "z_loss": 32.627941608428955, "grad_norm": 0.15910430252552032, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148231.311989998, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 24939.967148578726}
{"step": 2610, "total_steps": 57221, "tokens": 1368391680, "source": "stack_v3", "source_tokens": {"stack_v3": 1026555904, "fineweb_edu": 341835776}, "loss": 1.0392305999994278, "ce_loss": 1.0361029207706451, "perplexity": 2.81821278758834, "z_loss": 31.276761054992676, "grad_norm": 0.11838750541210175, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 858304.8372058058, "data_wait_seconds": 0.40147411823272705, "eta_seconds": 33358.321052004816}
{"step": 2620, "total_steps": 57221, "tokens": 1373634560, "source": "stack_v3", "source_tokens": {"stack_v3": 1030225920, "fineweb_edu": 343408640}, "loss": 1.1246391087770462, "ce_loss": 1.121852159500122, "perplexity": 3.070536062352439, "z_loss": 27.869472980499268, "grad_norm": 0.23669210076332092, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166055.6236862792, "data_wait_seconds": 2.1457672119140625e-05, "eta_seconds": 24549.742618197575}
{"step": 2630, "total_steps": 57221, "tokens": 1378877440, "source": "stack_v3", "source_tokens": {"stack_v3": 1034420224, "fineweb_edu": 344457216}, "loss": 0.953469768166542, "ce_loss": 0.950592502951622, "perplexity": 2.5872421538799717, "z_loss": 28.772750854492188, "grad_norm": 0.11677860468626022, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168064.5786553037, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 24503.030982198892}
{"step": 2640, "total_steps": 57221, "tokens": 1384120320, "source": "stack_v3", "source_tokens": {"stack_v3": 1038090240, "fineweb_edu": 346030080}, "loss": 0.8131489902734756, "ce_loss": 0.809998095035553, "perplexity": 2.2479037044957555, "z_loss": 31.509116649627686, "grad_norm": 0.12959417700767517, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130924.8756777507, "data_wait_seconds": 1.710653305053711e-05, "eta_seconds": 25303.077415156178}
{"step": 2650, "total_steps": 57221, "tokens": 1389363200, "source": "stack_v3", "source_tokens": {"stack_v3": 1042284544, "fineweb_edu": 347078656}, "loss": 0.7775034233927727, "ce_loss": 0.774287074804306, "perplexity": 2.169045208850446, "z_loss": 32.16339302062988, "grad_norm": 0.08674152195453644, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164978.4047137739, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 24558.94176598871}
{"step": 2660, "total_steps": 57221, "tokens": 1394606080, "source": "stack_v3", "source_tokens": {"stack_v3": 1045954560, "fineweb_edu": 348651520}, "loss": 1.0101557299494743, "ce_loss": 1.007251262664795, "perplexity": 2.7380644417747377, "z_loss": 29.044628143310547, "grad_norm": 0.156256765127182, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164088.259906027, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 24573.217431390658}
{"step": 2670, "total_steps": 57221, "tokens": 1399848960, "source": "stack_v3", "source_tokens": {"stack_v3": 1050148864, "fineweb_edu": 349700096}, "loss": 0.6983232945203781, "ce_loss": 0.695436991751194, "perplexity": 2.004584869622087, "z_loss": 28.863003492355347, "grad_norm": 0.11939216405153275, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129406.0458135456, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 25323.178626512876}
{"step": 2680, "total_steps": 57221, "tokens": 1405091840, "source": "stack_v3", "source_tokens": {"stack_v3": 1053818880, "fineweb_edu": 351272960}, "loss": 0.911052793264389, "ce_loss": 0.9082115292549133, "perplexity": 2.4798833660562183, "z_loss": 28.41264533996582, "grad_norm": 0.15197370946407318, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1113321.8847616198, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 25684.31336111086}
{"step": 2690, "total_steps": 57221, "tokens": 1410334720, "source": "stack_v3", "source_tokens": {"stack_v3": 1058013184, "fineweb_edu": 352321536}, "loss": 0.6382263973355293, "ce_loss": 0.6354272440075874, "perplexity": 1.8878285326114883, "z_loss": 27.991450309753418, "grad_norm": 0.09517205506563187, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1170384.6195268333, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 24427.581158370245}
{"step": 2700, "total_steps": 57221, "tokens": 1415577600, "source": "stack_v3", "source_tokens": {"stack_v3": 1061683200, "fineweb_edu": 353894400}, "loss": 0.9048052057623863, "ce_loss": 0.901746079325676, "perplexity": 2.4639015248923064, "z_loss": 30.59143304824829, "grad_norm": 0.14803935587406158, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136099.190976771, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 25160.146778578634}
{"step": 2710, "total_steps": 57221, "tokens": 1420820480, "source": "stack_v3", "source_tokens": {"stack_v3": 1065877504, "fineweb_edu": 354942976}, "loss": 0.9223488867282867, "ce_loss": 0.9198571518063545, "perplexity": 2.5089319679373405, "z_loss": 24.917428016662598, "grad_norm": 0.1439880132675171, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 893931.3700298511, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 31970.216593971476}
{"step": 2720, "total_steps": 57221, "tokens": 1426063360, "source": "stack_v3", "source_tokens": {"stack_v3": 1069547520, "fineweb_edu": 356515840}, "loss": 0.7782892882823944, "ce_loss": 0.7754042148590088, "perplexity": 2.171469690124139, "z_loss": 28.850828170776367, "grad_norm": 0.105084627866745, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142470.2218218336, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 25010.66206735326}
{"step": 2730, "total_steps": 57221, "tokens": 1431306240, "source": "stack_v3", "source_tokens": {"stack_v3": 1073741824, "fineweb_edu": 357564416}, "loss": 0.9286243841052055, "ce_loss": 0.9254240021109581, "perplexity": 2.522937764544092, "z_loss": 32.00385570526123, "grad_norm": 0.15935027599334717, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1175760.8348915074, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 24298.048473979114}
{"step": 2740, "total_steps": 57221, "tokens": 1436549120, "source": "stack_v3", "source_tokens": {"stack_v3": 1077411840, "fineweb_edu": 359137280}, "loss": 0.7321867942810059, "ce_loss": 0.7297303825616837, "perplexity": 2.0745212051721054, "z_loss": 24.564287185668945, "grad_norm": 0.15056276321411133, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150567.0632245874, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 24825.54193751025}
{"step": 2750, "total_steps": 57221, "tokens": 1441792000, "source": "stack_v3", "source_tokens": {"stack_v3": 1081606144, "fineweb_edu": 360185856}, "loss": 0.9126530438661575, "ce_loss": 0.9093644469976425, "perplexity": 2.4827441163768644, "z_loss": 32.88588094711304, "grad_norm": 0.18244454264640808, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 466607.04504936223, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 61203.979457658716}
{"step": 2760, "total_steps": 57221, "tokens": 1447034880, "source": "stack_v3", "source_tokens": {"stack_v3": 1085276160, "fineweb_edu": 361758720}, "loss": 0.965469092130661, "ce_loss": 0.9631198197603226, "perplexity": 2.619857219097444, "z_loss": 23.492828369140625, "grad_norm": 0.31125861406326294, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145280.5905439677, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 24930.977924316656}
{"step": 2770, "total_steps": 57221, "tokens": 1452277760, "source": "stack_v3", "source_tokens": {"stack_v3": 1089470464, "fineweb_edu": 362807296}, "loss": 1.0683235675096512, "ce_loss": 1.0658590197563171, "perplexity": 2.9033319326495826, "z_loss": 24.645302295684814, "grad_norm": 0.16740316152572632, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 815521.6325865795, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 35005.47514534416}
{"step": 2780, "total_steps": 57221, "tokens": 1457520640, "source": "stack_v3", "source_tokens": {"stack_v3": 1093140480, "fineweb_edu": 364380160}, "loss": 0.6308239512145519, "ce_loss": 0.6278178542852402, "perplexity": 1.8735178265003252, "z_loss": 30.06107473373413, "grad_norm": 0.13735701143741608, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1167204.0342645866, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 24453.71890612389}
{"step": 2790, "total_steps": 57221, "tokens": 1462763520, "source": "stack_v3", "source_tokens": {"stack_v3": 1097334784, "fineweb_edu": 365428736}, "loss": 0.8304260745644569, "ce_loss": 0.8276556879281998, "perplexity": 2.2879487823448343, "z_loss": 27.70388698577881, "grad_norm": 0.1301194578409195, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1104210.4296277163, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 25844.02004753868}
{"step": 2800, "total_steps": 57221, "tokens": 1468006400, "source": "stack_v3", "source_tokens": {"stack_v3": 1101004800, "fineweb_edu": 367001600}, "loss": 0.7341203913092613, "ce_loss": 0.7314706146717072, "perplexity": 2.0781344966573263, "z_loss": 26.497809886932373, "grad_norm": 0.11770729720592499, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1182879.6233157255, "data_wait_seconds": 2.467632293701172e-05, "eta_seconds": 24120.792207090417}
{"step": 2810, "total_steps": 57221, "tokens": 1473249280, "source": "stack_v3", "source_tokens": {"stack_v3": 1105199104, "fineweb_edu": 368050176}, "loss": 0.8087502494454384, "ce_loss": 0.8058023601770401, "perplexity": 2.2384918551836717, "z_loss": 29.478873252868652, "grad_norm": 0.09991463273763657, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140761.0925792279, "data_wait_seconds": 2.4080276489257812e-05, "eta_seconds": 25006.770397035405}
{"step": 2820, "total_steps": 57221, "tokens": 1478492160, "source": "stack_v3", "source_tokens": {"stack_v3": 1108869120, "fineweb_edu": 369623040}, "loss": 0.7364117056131363, "ce_loss": 0.7335285358130932, "perplexity": 2.082415537083577, "z_loss": 28.83174991607666, "grad_norm": 0.18039697408676147, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122374.3368999348, "data_wait_seconds": 2.187490463256836e-05, "eta_seconds": 25411.760499423137}
{"step": 2830, "total_steps": 57221, "tokens": 1483735040, "source": "stack_v3", "source_tokens": {"stack_v3": 1113063424, "fineweb_edu": 370671616}, "loss": 0.9291311651468277, "ce_loss": 0.9265821129083633, "perplexity": 2.5258612985717948, "z_loss": 25.490562438964844, "grad_norm": 0.1419304609298706, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1176753.7434992937, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 24232.992771454155}
{"step": 2840, "total_steps": 57221, "tokens": 1488977920, "source": "stack_v3", "source_tokens": {"stack_v3": 1116733440, "fineweb_edu": 372244480}, "loss": 0.6859617233276367, "ce_loss": 0.6834814101457596, "perplexity": 1.9807615860002865, "z_loss": 24.80311918258667, "grad_norm": 0.36101046204566956, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 918454.8351382653, "data_wait_seconds": 1.633167266845703e-05, "eta_seconds": 31042.37790387147}
{"step": 2850, "total_steps": 57221, "tokens": 1494220800, "source": "stack_v3", "source_tokens": {"stack_v3": 1120927744, "fineweb_edu": 373293056}, "loss": 0.5308474078774452, "ce_loss": 0.5284370630979538, "perplexity": 1.6962790585906287, "z_loss": 24.103497982025146, "grad_norm": 0.09441673010587692, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145504.134242705, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 24884.92040130892}
{"step": 2860, "total_steps": 57221, "tokens": 1499463680, "source": "stack_v3", "source_tokens": {"stack_v3": 1124597760, "fineweb_edu": 374865920}, "loss": 0.4391394220292568, "ce_loss": 0.43618031591176987, "perplexity": 1.5467876800357345, "z_loss": 29.591023921966553, "grad_norm": 0.1441839635372162, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155667.1106147242, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 24661.5448845299}
{"step": 2870, "total_steps": 57221, "tokens": 1504706560, "source": "stack_v3", "source_tokens": {"stack_v3": 1128792064, "fineweb_edu": 375914496}, "loss": 0.472026102244854, "ce_loss": 0.4696220336481929, "perplexity": 1.5993895635212756, "z_loss": 24.040642738342285, "grad_norm": 0.14769482612609863, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1185006.435180793, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24046.5305453405}
{"step": 2880, "total_steps": 57221, "tokens": 1509949440, "source": "stack_v3", "source_tokens": {"stack_v3": 1132462080, "fineweb_edu": 377487360}, "loss": 0.40372880175709724, "ce_loss": 0.4012788310647011, "perplexity": 1.4937337097995598, "z_loss": 24.499696731567383, "grad_norm": 0.13987413048744202, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118451.5989986022, "data_wait_seconds": 2.187490463256836e-05, "eta_seconds": 25472.761258071754}
{"step": 2890, "total_steps": 57221, "tokens": 1515192320, "source": "stack_v3", "source_tokens": {"stack_v3": 1136656384, "fineweb_edu": 378535936}, "loss": 0.3753066882491112, "ce_loss": 0.3727621268481016, "perplexity": 1.451738969032886, "z_loss": 25.445693969726562, "grad_norm": 0.08361177891492844, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1163175.2940329527, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 24488.83485243027}
{"step": 2900, "total_steps": 57221, "tokens": 1520435200, "source": "stack_v3", "source_tokens": {"stack_v3": 1140326400, "fineweb_edu": 380108800}, "loss": 0.7553363516926765, "ce_loss": 0.7525937780737877, "perplexity": 2.1224981722532585, "z_loss": 27.42586898803711, "grad_norm": 0.6699129939079285, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142792.256051772, "data_wait_seconds": 2.181529998779297e-05, "eta_seconds": 24921.034115504004}
{"step": 2910, "total_steps": 57221, "tokens": 1525678080, "source": "stack_v3", "source_tokens": {"stack_v3": 1144520704, "fineweb_edu": 381157376}, "loss": 1.2657648622989655, "ce_loss": 1.2629643082618713, "perplexity": 3.5358874278801347, "z_loss": 28.005656242370605, "grad_norm": 0.16234152019023895, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 448766.4360605711, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 63450.20400803047}
{"step": 2920, "total_steps": 57221, "tokens": 1530920960, "source": "stack_v3", "source_tokens": {"stack_v3": 1148190720, "fineweb_edu": 382730240}, "loss": 0.7509011849761009, "ce_loss": 0.7483231425285339, "perplexity": 2.1134530839990715, "z_loss": 25.780470848083496, "grad_norm": 0.2576635479927063, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122599.0560935733, "data_wait_seconds": 2.3603439331054688e-05, "eta_seconds": 25359.970583858198}
{"step": 2930, "total_steps": 57221, "tokens": 1536163840, "source": "stack_v3", "source_tokens": {"stack_v3": 1152385024, "fineweb_edu": 383778816}, "loss": 0.682296022772789, "ce_loss": 0.6801045686006546, "perplexity": 1.9740841486549205, "z_loss": 21.91460418701172, "grad_norm": 0.1345396786928177, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148915.6918760869, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 24774.521195302717}
{"step": 2940, "total_steps": 57221, "tokens": 1541406720, "source": "stack_v3", "source_tokens": {"stack_v3": 1156055040, "fineweb_edu": 385351680}, "loss": 0.6441790089011192, "ce_loss": 0.6416294761002064, "perplexity": 1.89957366870115, "z_loss": 25.495408535003662, "grad_norm": 0.1276664286851883, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1003928.9572561793, "data_wait_seconds": 0.1898919939994812, "eta_seconds": 28347.218271081336}
{"step": 2950, "total_steps": 57221, "tokens": 1546649600, "source": "stack_v3", "source_tokens": {"stack_v3": 1160249344, "fineweb_edu": 386400256}, "loss": 0.804619237780571, "ce_loss": 0.8018373250961304, "perplexity": 2.229633729444443, "z_loss": 27.81912612915039, "grad_norm": 0.10360319167375565, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1089721.407195544, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 26110.664810399758}
{"step": 2960, "total_steps": 57221, "tokens": 1551892480, "source": "stack_v3", "source_tokens": {"stack_v3": 1163919360, "fineweb_edu": 387973120}, "loss": 0.5623477548360825, "ce_loss": 0.5597100183367729, "perplexity": 1.7501649109720545, "z_loss": 26.377201080322266, "grad_norm": 0.1417350471019745, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1184694.9246471422, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 24013.02388332016}
{"step": 2970, "total_steps": 57221, "tokens": 1557135360, "source": "stack_v3", "source_tokens": {"stack_v3": 1168113664, "fineweb_edu": 389021696}, "loss": 0.8957558050751686, "ce_loss": 0.8931665793061256, "perplexity": 2.4428529041441496, "z_loss": 25.892228603363037, "grad_norm": 0.2827361822128296, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157398.9964535753, "data_wait_seconds": 1.537799835205078e-05, "eta_seconds": 24574.813635706203}
{"step": 2980, "total_steps": 57221, "tokens": 1562378240, "source": "stack_v3", "source_tokens": {"stack_v3": 1171783680, "fineweb_edu": 390594560}, "loss": 0.9440176412463188, "ce_loss": 0.9415073469281197, "perplexity": 2.5638431081782223, "z_loss": 25.103126764297485, "grad_norm": 0.12391319870948792, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168012.4014898986, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 24347.020394411404}
{"step": 2990, "total_steps": 57221, "tokens": 1567621120, "source": "stack_v3", "source_tokens": {"stack_v3": 1175977984, "fineweb_edu": 391643136}, "loss": 0.9388905316591263, "ce_loss": 0.9366055577993393, "perplexity": 2.551306441054058, "z_loss": 22.849724292755127, "grad_norm": 0.146034374833107, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135746.952858788, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 25034.078945519403}
{"step": 3000, "total_steps": 57221, "tokens": 1572864000, "source": "stack_v3", "source_tokens": {"stack_v3": 1179648000, "fineweb_edu": 393216000}, "loss": 0.9098236188292503, "ce_loss": 0.9072393551468849, "perplexity": 2.4774736591739166, "z_loss": 25.842592239379883, "grad_norm": 0.1330440640449524, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164686.3089378534, "data_wait_seconds": 1.609325408935547e-05, "eta_seconds": 24407.547149690796}
{"step": 3010, "total_steps": 57221, "tokens": 1578106880, "source": "stack_v3", "source_tokens": {"stack_v3": 1183842304, "fineweb_edu": 394264576}, "loss": 0.9003111571073532, "ce_loss": 0.8980232775211334, "perplexity": 2.454745960598963, "z_loss": 22.87885808944702, "grad_norm": 0.12140854448080063, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1102089.8567869675, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 25789.088743508793}
{"step": 3020, "total_steps": 57221, "tokens": 1583349760, "source": "stack_v3", "source_tokens": {"stack_v3": 1187512320, "fineweb_edu": 395837440}, "loss": 1.0284964591264725, "ce_loss": 1.0260759592056274, "perplexity": 2.790095875456158, "z_loss": 24.205244064331055, "grad_norm": 0.17355087399482727, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147278.511227011, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 24768.746177951572}
{"step": 3030, "total_steps": 57221, "tokens": 1588592640, "source": "stack_v3", "source_tokens": {"stack_v3": 1191706624, "fineweb_edu": 396886016}, "loss": 0.7970011979341507, "ce_loss": 0.7945060208439827, "perplexity": 2.213347379243801, "z_loss": 24.951945781707764, "grad_norm": 0.13024306297302246, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152071.490939203, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 24661.14957574219}
{"step": 3040, "total_steps": 57221, "tokens": 1593835520, "source": "stack_v3", "source_tokens": {"stack_v3": 1195376640, "fineweb_edu": 398458880}, "loss": 0.8485667556524277, "ce_loss": 0.8461313731968403, "perplexity": 2.3306131167465023, "z_loss": 24.35380983352661, "grad_norm": 0.15076524019241333, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1181854.186292324, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 24035.25308745145}
{"step": 3050, "total_steps": 57221, "tokens": 1599078400, "source": "stack_v3", "source_tokens": {"stack_v3": 1199570944, "fineweb_edu": 399507456}, "loss": 0.7810973115265369, "ce_loss": 0.778570581227541, "perplexity": 2.178356255662339, "z_loss": 25.267611026763916, "grad_norm": 0.09727637469768524, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1112306.4330601695, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 25533.360911941854}
{"step": 3060, "total_steps": 57221, "tokens": 1604321280, "source": "stack_v3", "source_tokens": {"stack_v3": 1203240960, "fineweb_edu": 401080320}, "loss": 1.0090707913041115, "ce_loss": 1.0060702487826347, "perplexity": 2.734832658424892, "z_loss": 30.00527858734131, "grad_norm": 0.1398565173149109, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 665031.9138006786, "data_wait_seconds": 0.8462470173835754, "eta_seconds": 42698.219635382295}
{"step": 3070, "total_steps": 57221, "tokens": 1609564160, "source": "stack_v3", "source_tokens": {"stack_v3": 1207435264, "fineweb_edu": 402128896}, "loss": 0.7095839232206345, "ce_loss": 0.7067093104124069, "perplexity": 2.0273090253983086, "z_loss": 28.746041297912598, "grad_norm": 0.11431810259819031, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132208.5074308184, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 25075.27160736756}
{"step": 3080, "total_steps": 57221, "tokens": 1614807040, "source": "stack_v3", "source_tokens": {"stack_v3": 1211105280, "fineweb_edu": 403701760}, "loss": 1.1859364956617355, "ce_loss": 1.1835522949695587, "perplexity": 3.2659552574823953, "z_loss": 23.842082500457764, "grad_norm": 0.2573206126689911, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149963.1563608083, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 24683.56729778042}
{"step": 3090, "total_steps": 57221, "tokens": 1620049920, "source": "stack_v3", "source_tokens": {"stack_v3": 1215299584, "fineweb_edu": 404750336}, "loss": 1.0322748720645905, "ce_loss": 1.0297814309597015, "perplexity": 2.800453675329756, "z_loss": 24.93415069580078, "grad_norm": 0.08489779382944107, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1106613.0403089225, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 25645.775936344868}
{"step": 3100, "total_steps": 57221, "tokens": 1625292800, "source": "stack_v3", "source_tokens": {"stack_v3": 1218969600, "fineweb_edu": 406323200}, "loss": 0.6867280155420303, "ce_loss": 0.6851427927613258, "perplexity": 1.9840551240206878, "z_loss": 15.852173328399658, "grad_norm": 0.22552813589572906, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155770.0648165187, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 24550.47769774566}
{"step": 3110, "total_steps": 57221, "tokens": 1630535680, "source": "stack_v3", "source_tokens": {"stack_v3": 1223163904, "fineweb_edu": 407371776}, "loss": 0.965177409350872, "ce_loss": 0.9626223370432854, "perplexity": 2.6185542095490764, "z_loss": 25.550467014312744, "grad_norm": 0.12074550986289978, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1161049.8839896927, "data_wait_seconds": 1.8596649169921875e-05, "eta_seconds": 24434.31992991944}
{"step": 3120, "total_steps": 57221, "tokens": 1635778560, "source": "stack_v3", "source_tokens": {"stack_v3": 1226833920, "fineweb_edu": 408944640}, "loss": 1.1386842131614685, "ce_loss": 1.1361577212810516, "perplexity": 3.114777500511464, "z_loss": 25.265122413635254, "grad_norm": 0.18405458331108093, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143987.4577600495, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 24794.17169095343}
{"step": 3130, "total_steps": 57221, "tokens": 1641021440, "source": "stack_v3", "source_tokens": {"stack_v3": 1231028224, "fineweb_edu": 409993216}, "loss": 0.8726909458637238, "ce_loss": 0.8697109818458557, "perplexity": 2.3862210926367946, "z_loss": 29.799564361572266, "grad_norm": 0.4246808588504791, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1068661.7670502355, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 26536.90759264049}
{"step": 3140, "total_steps": 57221, "tokens": 1646264320, "source": "stack_v3", "source_tokens": {"stack_v3": 1234698240, "fineweb_edu": 411566080}, "loss": 0.8439227417111397, "ce_loss": 0.8406258821487427, "perplexity": 2.317817203309903, "z_loss": 32.96872138977051, "grad_norm": 0.1055682897567749, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151960.927172585, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24613.45260172361}
{"step": 3150, "total_steps": 57221, "tokens": 1651507200, "source": "stack_v3", "source_tokens": {"stack_v3": 1238892544, "fineweb_edu": 412614656}, "loss": 0.8636949993669987, "ce_loss": 0.8616641461849213, "perplexity": 2.3670966126224675, "z_loss": 20.308634996414185, "grad_norm": 0.13882815837860107, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130059.3352227753, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 25085.844536129152}
{"step": 3160, "total_steps": 57221, "tokens": 1656750080, "source": "stack_v3", "source_tokens": {"stack_v3": 1242562560, "fineweb_edu": 414187520}, "loss": 1.3074502646923065, "ce_loss": 1.304345890879631, "perplexity": 3.685277731368764, "z_loss": 31.04348659515381, "grad_norm": 0.34888792037963867, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 544456.4828265287, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 52057.879397186916}
{"step": 3170, "total_steps": 57221, "tokens": 1661992960, "source": "stack_v3", "source_tokens": {"stack_v3": 1246756864, "fineweb_edu": 415236096}, "loss": 0.9052572920918465, "ce_loss": 0.9027548879384995, "perplexity": 2.466388384143377, "z_loss": 25.024038791656494, "grad_norm": 0.09072062373161316, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1186497.2406553, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 23883.75300759147}
{"step": 3180, "total_steps": 57221, "tokens": 1667235840, "source": "stack_v3", "source_tokens": {"stack_v3": 1250426880, "fineweb_edu": 416808960}, "loss": 0.9797549769282341, "ce_loss": 0.9778038263320923, "perplexity": 2.658611054164691, "z_loss": 19.511695384979248, "grad_norm": 0.185980424284935, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150747.26903297, "data_wait_seconds": 1.704692840576172e-05, "eta_seconds": 24621.187399218794}
{"step": 3190, "total_steps": 57221, "tokens": 1672478720, "source": "stack_v3", "source_tokens": {"stack_v3": 1254621184, "fineweb_edu": 417857536}, "loss": 0.8700532391667366, "ce_loss": 0.8674573972821236, "perplexity": 2.380849596450547, "z_loss": 25.95846652984619, "grad_norm": 0.1983753889799118, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141282.8067599765, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 24820.772828795947}
{"step": 3200, "total_steps": 57221, "tokens": 1677721600, "source": "stack_v3", "source_tokens": {"stack_v3": 1258291200, "fineweb_edu": 419430400}, "loss": 0.8183083608746529, "ce_loss": 0.8159697130322456, "perplexity": 2.2613674869851645, "z_loss": 23.3863263130188, "grad_norm": 0.2754775881767273, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155770.0648165187, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 24505.115041629175}
{"step": 3210, "total_steps": 57221, "tokens": 1682964480, "source": "stack_v3", "source_tokens": {"stack_v3": 1262485504, "fineweb_edu": 420478976}, "loss": 0.8493962734937668, "ce_loss": 0.8476097583770752, "perplexity": 2.3340612088152497, "z_loss": 17.86538815498352, "grad_norm": 0.1730155050754547, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142558.668157502, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 24783.87877067551}
{"step": 3220, "total_steps": 57221, "tokens": 1688207360, "source": "stack_v3", "source_tokens": {"stack_v3": 1266155520, "fineweb_edu": 422051840}, "loss": 0.7011999636888504, "ce_loss": 0.6992317698895931, "perplexity": 2.0122062760893153, "z_loss": 19.681789875030518, "grad_norm": 0.1200409084558487, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1139334.2066333522, "data_wait_seconds": 2.1278858184814453e-05, "eta_seconds": 24849.41861234838}
{"step": 3230, "total_steps": 57221, "tokens": 1693450240, "source": "stack_v3", "source_tokens": {"stack_v3": 1270349824, "fineweb_edu": 423100416}, "loss": 0.7321712002158165, "ce_loss": 0.7298502996563911, "perplexity": 2.0747699906444557, "z_loss": 23.20901918411255, "grad_norm": 0.1347450315952301, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141952.9960458414, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 24787.841406796128}
{"step": 3240, "total_steps": 57221, "tokens": 1698693120, "source": "stack_v3", "source_tokens": {"stack_v3": 1274019840, "fineweb_edu": 424673280}, "loss": 0.6989771202206612, "ce_loss": 0.6971292495727539, "perplexity": 2.0079800159679317, "z_loss": 18.47888469696045, "grad_norm": 0.13782452046871185, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177054.633412379, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 24044.174396520713}
{"step": 3250, "total_steps": 57221, "tokens": 1703936000, "source": "stack_v3", "source_tokens": {"stack_v3": 1278214144, "fineweb_edu": 425721856}, "loss": 0.6438993960618973, "ce_loss": 0.6419670134782791, "perplexity": 1.9002149540395368, "z_loss": 19.323785305023193, "grad_norm": 0.09084321558475494, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144965.3197639803, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 24713.468182453653}
{"step": 3260, "total_steps": 57221, "tokens": 1709178880, "source": "stack_v3", "source_tokens": {"stack_v3": 1281884160, "fineweb_edu": 427294720}, "loss": 0.7946926653385162, "ce_loss": 0.7925929427146912, "perplexity": 2.2091171204779902, "z_loss": 20.99717664718628, "grad_norm": 0.2078254669904709, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1107250.7569416435, "data_wait_seconds": 2.1576881408691406e-05, "eta_seconds": 25550.509622718673}
{"step": 3270, "total_steps": 57221, "tokens": 1714421760, "source": "stack_v3", "source_tokens": {"stack_v3": 1286078464, "fineweb_edu": 428343296}, "loss": 0.8556467071175575, "ce_loss": 0.853149063885212, "perplexity": 2.347026162193131, "z_loss": 24.976505279541016, "grad_norm": 0.10113415122032166, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1174830.7308278093, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24076.300949388184}
{"step": 3280, "total_steps": 57221, "tokens": 1719664640, "source": "stack_v3", "source_tokens": {"stack_v3": 1289748480, "fineweb_edu": 429916160}, "loss": 0.9791994616389275, "ce_loss": 0.9773368239402771, "perplexity": 2.6573697663082485, "z_loss": 18.626288652420044, "grad_norm": 0.18730567395687103, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1015494.7226705853, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 27848.825531685026}
{"step": 3290, "total_steps": 57221, "tokens": 1724907520, "source": "stack_v3", "source_tokens": {"stack_v3": 1293942784, "fineweb_edu": 430964736}, "loss": 0.5982586890459061, "ce_loss": 0.5963625460863113, "perplexity": 1.8155029669180465, "z_loss": 18.96157217025757, "grad_norm": 0.1919364333152771, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098677.6215934483, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 25735.5678538275}
{"step": 3300, "total_steps": 57221, "tokens": 1730150400, "source": "stack_v3", "source_tokens": {"stack_v3": 1297612800, "fineweb_edu": 432537600}, "loss": 0.7271003350615501, "ce_loss": 0.7252122461795807, "perplexity": 2.0651693777640645, "z_loss": 18.88079023361206, "grad_norm": 0.261689692735672, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1101415.203672507, "data_wait_seconds": 2.1457672119140625e-05, "eta_seconds": 25666.841628605038}
{"step": 3310, "total_steps": 57221, "tokens": 1735393280, "source": "stack_v3", "source_tokens": {"stack_v3": 1301807104, "fineweb_edu": 433586176}, "loss": 0.7065043151378632, "ce_loss": 0.7039058730006218, "perplexity": 2.0216335505675787, "z_loss": 25.984479904174805, "grad_norm": 0.14296285808086395, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1187549.2238008701, "data_wait_seconds": 1.6391277313232422e-05, "eta_seconds": 23800.787498758407}
{"step": 3320, "total_steps": 57221, "tokens": 1740636160, "source": "stack_v3", "source_tokens": {"stack_v3": 1305477120, "fineweb_edu": 435159040}, "loss": 0.5350096151232719, "ce_loss": 0.5326963439583778, "perplexity": 1.7035193958879964, "z_loss": 23.132672786712646, "grad_norm": 0.13598278164863586, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1070786.44296053, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 26391.223035909934}
{"step": 3330, "total_steps": 57221, "tokens": 1745879040, "source": "stack_v3", "source_tokens": {"stack_v3": 1309671424, "fineweb_edu": 436207616}, "loss": 0.8381363227963448, "ce_loss": 0.8359192609786987, "perplexity": 2.306933748034122, "z_loss": 22.17036199569702, "grad_norm": 0.10796685516834259, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1162835.0918196396, "data_wait_seconds": 1.7762184143066406e-05, "eta_seconds": 24297.616367757786}
{"step": 3340, "total_steps": 57221, "tokens": 1751121920, "source": "stack_v3", "source_tokens": {"stack_v3": 1313341440, "fineweb_edu": 437780480}, "loss": 0.6119535714387894, "ce_loss": 0.6099679991602898, "perplexity": 1.8403725043737877, "z_loss": 19.85582971572876, "grad_norm": 0.12259340286254883, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152792.8486457712, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24504.730501394948}
{"step": 3350, "total_steps": 57221, "tokens": 1756364800, "source": "stack_v3", "source_tokens": {"stack_v3": 1317535744, "fineweb_edu": 438829056}, "loss": 0.7675188407301903, "ce_loss": 0.7650339528918266, "perplexity": 2.149067340467452, "z_loss": 24.849059104919434, "grad_norm": 0.24949687719345093, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1081143.1143770455, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 26123.86355184251}
{"step": 3360, "total_steps": 57221, "tokens": 1761607680, "source": "stack_v3", "source_tokens": {"stack_v3": 1321205760, "fineweb_edu": 440401920}, "loss": 0.8570749014616013, "ce_loss": 0.8551993072032928, "perplexity": 2.351843073131939, "z_loss": 18.75600028038025, "grad_norm": 0.12393105775117874, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136788.5146761516, "data_wait_seconds": 1.7464160919189453e-05, "eta_seconds": 24840.49755555857}
{"step": 3370, "total_steps": 57221, "tokens": 1766850560, "source": "stack_v3", "source_tokens": {"stack_v3": 1325400064, "fineweb_edu": 441450496}, "loss": 1.0141642838716507, "ce_loss": 1.0121306329965591, "perplexity": 2.7514571195242823, "z_loss": 20.336644172668457, "grad_norm": 0.11490330845117569, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131422.4358706616, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 24953.676491551796}
{"step": 3380, "total_steps": 57221, "tokens": 1772093440, "source": "stack_v3", "source_tokens": {"stack_v3": 1329070080, "fineweb_edu": 443023360}, "loss": 0.984184518456459, "ce_loss": 0.9814665615558624, "perplexity": 2.6683666977823335, "z_loss": 27.179826736450195, "grad_norm": 0.1713837832212448, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 869151.6349826236, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 32477.539503868436}
{"step": 3390, "total_steps": 57221, "tokens": 1777336320, "source": "stack_v3", "source_tokens": {"stack_v3": 1333264384, "fineweb_edu": 444071936}, "loss": 0.8600956350564957, "ce_loss": 0.8582076281309128, "perplexity": 2.3589288246307465, "z_loss": 18.879866361618042, "grad_norm": 0.1089874878525734, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1179533.9219704992, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 23926.962297830265}
{"step": 3400, "total_steps": 57221, "tokens": 1782579200, "source": "stack_v3", "source_tokens": {"stack_v3": 1336934400, "fineweb_edu": 445644800}, "loss": 1.085546687245369, "ce_loss": 1.0833915770053864, "perplexity": 2.954683613703997, "z_loss": 21.551174640655518, "grad_norm": 0.14876815676689148, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108144.2376086067, "data_wait_seconds": 2.1517276763916016e-05, "eta_seconds": 25463.67146292585}
{"step": 3410, "total_steps": 57221, "tokens": 1787822080, "source": "stack_v3", "source_tokens": {"stack_v3": 1341128704, "fineweb_edu": 446693376}, "loss": 1.3784914910793304, "ce_loss": 1.3756579011678696, "perplexity": 3.9576796286490743, "z_loss": 28.335937023162842, "grad_norm": 0.1325022429227829, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1064836.5048009625, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 26494.37523300666}
{"step": 3420, "total_steps": 57221, "tokens": 1793064960, "source": "stack_v3", "source_tokens": {"stack_v3": 1344798720, "fineweb_edu": 448266240}, "loss": 1.0496081560850143, "ce_loss": 1.0465264916419983, "perplexity": 2.847742262229838, "z_loss": 30.816749811172485, "grad_norm": 0.39974886178970337, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136954.5541745361, "data_wait_seconds": 2.1696090698242188e-05, "eta_seconds": 24809.201859857185}
{"step": 3430, "total_steps": 57221, "tokens": 1798307840, "source": "stack_v3", "source_tokens": {"stack_v3": 1348993024, "fineweb_edu": 449314816}, "loss": 1.0452445596456528, "ce_loss": 1.042806714773178, "perplexity": 2.8371689737630175, "z_loss": 24.3784499168396, "grad_norm": 0.08882538229227066, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1159287.5057183532, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 24326.74554059375}
{"step": 3440, "total_steps": 57221, "tokens": 1803550720, "source": "stack_v3", "source_tokens": {"stack_v3": 1352663040, "fineweb_edu": 450887680}, "loss": 0.8117525428533554, "ce_loss": 0.8098377883434296, "perplexity": 2.2475433793707116, "z_loss": 19.147574424743652, "grad_norm": 0.1853318065404892, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 977672.9059168058, "data_wait_seconds": 2.1517276763916016e-05, "eta_seconds": 28840.37095572264}
{"step": 3450, "total_steps": 57221, "tokens": 1808793600, "source": "stack_v3", "source_tokens": {"stack_v3": 1356857344, "fineweb_edu": 451936256}, "loss": 0.8608051538467407, "ce_loss": 0.8588081896305084, "perplexity": 2.360345931950562, "z_loss": 19.969878673553467, "grad_norm": 0.10541745275259018, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1172232.8510395128, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 24049.152329249773}
{"step": 3460, "total_steps": 57221, "tokens": 1814036480, "source": "stack_v3", "source_tokens": {"stack_v3": 1360527360, "fineweb_edu": 453509120}, "loss": 0.9118325859308243, "ce_loss": 0.9099761694669724, "perplexity": 2.484263331360044, "z_loss": 18.564115524291992, "grad_norm": 0.1748570203781128, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152882.960995745, "data_wait_seconds": 2.1755695343017578e-05, "eta_seconds": 24448.24364101607}
{"step": 3470, "total_steps": 57221, "tokens": 1819279360, "source": "stack_v3", "source_tokens": {"stack_v3": 1364721664, "fineweb_edu": 454557696}, "loss": 0.671687439084053, "ce_loss": 0.6695936843752861, "perplexity": 1.953443444771172, "z_loss": 20.9375581741333, "grad_norm": 0.12498295307159424, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146097.1533791162, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 24588.42215680657}
{"step": 3480, "total_steps": 57221, "tokens": 1824522240, "source": "stack_v3", "source_tokens": {"stack_v3": 1368391680, "fineweb_edu": 456130560}, "loss": 0.7822322249412537, "ce_loss": 0.7799652069807053, "perplexity": 2.1813963668119567, "z_loss": 22.67041254043579, "grad_norm": 0.2507317364215851, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1160899.5305939985, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 24270.38431618921}
{"step": 3490, "total_steps": 57221, "tokens": 1829765120, "source": "stack_v3", "source_tokens": {"stack_v3": 1372585984, "fineweb_edu": 457179136}, "loss": 0.9078660607337952, "ce_loss": 0.9060535132884979, "perplexity": 2.4745375084551022, "z_loss": 18.125428199768066, "grad_norm": 0.11963094770908356, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 884412.0132173756, "data_wait_seconds": 0.3552495837211609, "eta_seconds": 31851.936042252928}
{"step": 3500, "total_steps": 57221, "tokens": 1835008000, "source": "stack_v3", "source_tokens": {"stack_v3": 1376256000, "fineweb_edu": 458752000}, "loss": 0.9712686836719513, "ce_loss": 0.9688835442066193, "perplexity": 2.6350009544322015, "z_loss": 23.85154628753662, "grad_norm": 0.26179569959640503, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147447.0306962577, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 24545.788386336062}
{"step": 3510, "total_steps": 57221, "tokens": 1840250880, "source": "stack_v3", "source_tokens": {"stack_v3": 1380450304, "fineweb_edu": 459800576}, "loss": 0.9051380902528763, "ce_loss": 0.9027930796146393, "perplexity": 2.466482581448544, "z_loss": 23.449957847595215, "grad_norm": 0.14658941328525543, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122467.2045768, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 25087.3691500118}
{"step": 3520, "total_steps": 57221, "tokens": 1845493760, "source": "stack_v3", "source_tokens": {"stack_v3": 1384120320, "fineweb_edu": 461373440}, "loss": 0.8418806046247482, "ce_loss": 0.8395433872938156, "perplexity": 2.3153095356262137, "z_loss": 23.372308254241943, "grad_norm": 0.13369812071323395, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1104366.9217290357, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 25493.79711221368}
{"step": 3530, "total_steps": 57221, "tokens": 1850736640, "source": "stack_v3", "source_tokens": {"stack_v3": 1388314624, "fineweb_edu": 462422016}, "loss": 0.7988470643758774, "ce_loss": 0.7965873256325722, "perplexity": 2.217958826992507, "z_loss": 22.597370386123657, "grad_norm": 0.13261708617210388, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1181966.6876784689, "data_wait_seconds": 1.6987323760986328e-05, "eta_seconds": 23815.614816766698}
{"step": 3540, "total_steps": 57221, "tokens": 1855979520, "source": "stack_v3", "source_tokens": {"stack_v3": 1391984640, "fineweb_edu": 463994880}, "loss": 0.5877699870616198, "ce_loss": 0.585799677297473, "perplexity": 1.7964269732671132, "z_loss": 19.703069925308228, "grad_norm": 0.3770971894264221, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 858963.46258533, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 32765.096195467282}
{"step": 3550, "total_steps": 57221, "tokens": 1861222400, "source": "stack_v3", "source_tokens": {"stack_v3": 1396178944, "fineweb_edu": 465043456}, "loss": 0.7209730818867683, "ce_loss": 0.7194034606218338, "perplexity": 2.05320802580576, "z_loss": 15.69610869884491, "grad_norm": 0.19550028443336487, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138841.6965941328, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 24708.243194952374}
{"step": 3560, "total_steps": 57221, "tokens": 1866465280, "source": "stack_v3", "source_tokens": {"stack_v3": 1399848960, "fineweb_edu": 466616320}, "loss": 0.5672397958114743, "ce_loss": 0.5654014851897955, "perplexity": 1.7601543167707843, "z_loss": 18.38330340385437, "grad_norm": 0.10839033871889114, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166692.4631997044, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 24113.92514085723}
{"step": 3570, "total_steps": 57221, "tokens": 1871708160, "source": "stack_v3", "source_tokens": {"stack_v3": 1404043264, "fineweb_edu": 467664896}, "loss": 0.5992574521806091, "ce_loss": 0.5972967450506985, "perplexity": 1.8171998003762067, "z_loss": 19.606980562210083, "grad_norm": 0.16213414072990417, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144275.7524246394, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 24581.74245185056}
{"step": 3580, "total_steps": 57221, "tokens": 1876951040, "source": "stack_v3", "source_tokens": {"stack_v3": 1407713280, "fineweb_edu": 469237760}, "loss": 0.3393905535340309, "ce_loss": 0.33717119158245623, "perplexity": 1.4009788790020565, "z_loss": 22.193654894828796, "grad_norm": 0.1351637840270996, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138343.189061378, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 24705.246388120344}
{"step": 3590, "total_steps": 57221, "tokens": 1882193920, "source": "stack_v3", "source_tokens": {"stack_v3": 1411907584, "fineweb_edu": 470286336}, "loss": 0.6217701189452782, "ce_loss": 0.6198937623994425, "perplexity": 1.8587305642815244, "z_loss": 18.76363492012024, "grad_norm": 0.09264543652534485, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135133.244109397, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24770.489478581585}
{"step": 3600, "total_steps": 57221, "tokens": 1887436800, "source": "stack_v3", "source_tokens": {"stack_v3": 1415577600, "fineweb_edu": 471859200}, "loss": 0.8644513562321663, "ce_loss": 0.8622466251254082, "perplexity": 2.3684757981836957, "z_loss": 22.047272205352783, "grad_norm": 0.24996067583560944, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1079923.7547279927, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 26031.988903773017}
{"step": 3610, "total_steps": 57221, "tokens": 1892679680, "source": "stack_v3", "source_tokens": {"stack_v3": 1419771904, "fineweb_edu": 472907776}, "loss": 0.829479455947876, "ce_loss": 0.8271489441394806, "perplexity": 2.286789672221205, "z_loss": 23.30510187149048, "grad_norm": 0.14284534752368927, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119989.0027064332, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 25096.068132882705}
{"step": 3620, "total_steps": 57221, "tokens": 1897922560, "source": "stack_v3", "source_tokens": {"stack_v3": 1423441920, "fineweb_edu": 474480640}, "loss": 0.7325664833188057, "ce_loss": 0.7305703237652779, "perplexity": 2.076264413003615, "z_loss": 19.961485385894775, "grad_norm": 0.19601984322071075, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155368.4346657093, "data_wait_seconds": 2.4080276489257812e-05, "eta_seconds": 24323.04414490168}
{"step": 3630, "total_steps": 57221, "tokens": 1903165440, "source": "stack_v3", "source_tokens": {"stack_v3": 1427636224, "fineweb_edu": 475529216}, "loss": 0.901164636015892, "ce_loss": 0.899234265089035, "perplexity": 2.4577204280976837, "z_loss": 19.30369806289673, "grad_norm": 0.12759073078632355, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1097483.9164918896, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 25601.135595509782}
{"step": 3640, "total_steps": 57221, "tokens": 1908408320, "source": "stack_v3", "source_tokens": {"stack_v3": 1431306240, "fineweb_edu": 477102080}, "loss": 0.7383762001991272, "ce_loss": 0.7360108569264412, "perplexity": 2.087591182287067, "z_loss": 23.65348529815674, "grad_norm": 0.1631157398223877, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 996784.2853120924, "data_wait_seconds": 1.6033649444580078e-05, "eta_seconds": 28182.217651238898}
{"step": 3650, "total_steps": 57221, "tokens": 1913651200, "source": "stack_v3", "source_tokens": {"stack_v3": 1435500544, "fineweb_edu": 478150656}, "loss": 0.7616006061434746, "ce_loss": 0.7596199810504913, "perplexity": 2.1374637893931308, "z_loss": 19.806281089782715, "grad_norm": 0.09692392498254776, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149384.4536666023, "data_wait_seconds": 2.1576881408691406e-05, "eta_seconds": 24435.991552176416}
{"step": 3660, "total_steps": 57221, "tokens": 1918894080, "source": "stack_v3", "source_tokens": {"stack_v3": 1439170560, "fineweb_edu": 479723520}, "loss": 0.714185819029808, "ce_loss": 0.7126776799559593, "perplexity": 2.0394449346114767, "z_loss": 15.08143663406372, "grad_norm": 0.11318697780370712, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135385.1432227518, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 24732.669867682736}
{"step": 3670, "total_steps": 57221, "tokens": 1924136960, "source": "stack_v3", "source_tokens": {"stack_v3": 1443364864, "fineweb_edu": 480772096}, "loss": 1.0482828244566917, "ce_loss": 1.0464440509676933, "perplexity": 2.847507502114517, "z_loss": 18.387426137924194, "grad_norm": 0.17329034209251404, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1162119.9081306101, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 24159.1791376872}
{"step": 3680, "total_steps": 57221, "tokens": 1929379840, "source": "stack_v3", "source_tokens": {"stack_v3": 1447034880, "fineweb_edu": 482344960}, "loss": 1.2025507092475891, "ce_loss": 1.2006420642137527, "perplexity": 3.3222493354978164, "z_loss": 19.08638572692871, "grad_norm": 0.6365567445755005, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149341.3207365323, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 24423.223679116927}
{"step": 3690, "total_steps": 57221, "tokens": 1934622720, "source": "stack_v3", "source_tokens": {"stack_v3": 1451229184, "fineweb_edu": 483393536}, "loss": 0.8510980010032654, "ce_loss": 0.8493309915065765, "perplexity": 2.338082131774337, "z_loss": 17.670297622680664, "grad_norm": 0.12610240280628204, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 437781.7143422303, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 64108.15335713234}
{"step": 3700, "total_steps": 57221, "tokens": 1939865600, "source": "stack_v3", "source_tokens": {"stack_v3": 1454899200, "fineweb_edu": 484966400}, "loss": 0.9639807194471359, "ce_loss": 0.9621750861406326, "perplexity": 2.6173833206752812, "z_loss": 18.056496381759644, "grad_norm": 0.18924792110919952, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1103295.4346338485, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 25433.01958764321}
{"step": 3710, "total_steps": 57221, "tokens": 1945108480, "source": "stack_v3", "source_tokens": {"stack_v3": 1459093504, "fineweb_edu": 486014976}, "loss": 0.7748714834451675, "ce_loss": 0.7732129022479057, "perplexity": 2.1667165309383716, "z_loss": 16.585842847824097, "grad_norm": 0.13923253118991852, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1171680.3410379603, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 23944.15143566113}
{"step": 3720, "total_steps": 57221, "tokens": 1950351360, "source": "stack_v3", "source_tokens": {"stack_v3": 1462763520, "fineweb_edu": 487587840}, "loss": 0.7229844927787781, "ce_loss": 0.7210234999656677, "perplexity": 2.056536999394653, "z_loss": 19.609775066375732, "grad_norm": 0.12250965088605881, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1158724.9647325051, "data_wait_seconds": 1.5616416931152344e-05, "eta_seconds": 24207.3395272668}
{"step": 3730, "total_steps": 57221, "tokens": 1955594240, "source": "stack_v3", "source_tokens": {"stack_v3": 1466957824, "fineweb_edu": 488636416}, "loss": 0.70671246945858, "ce_loss": 0.7048308998346329, "perplexity": 2.0235044810474307, "z_loss": 18.815685749053955, "grad_norm": 0.11745250225067139, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177481.6357386904, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 23817.276557698846}
{"step": 3740, "total_steps": 57221, "tokens": 1960837120, "source": "stack_v3", "source_tokens": {"stack_v3": 1470627840, "fineweb_edu": 490209280}, "loss": 0.880271390080452, "ce_loss": 0.8785157054662704, "perplexity": 2.4073238756111657, "z_loss": 17.556952953338623, "grad_norm": 0.23953589797019958, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131987.2670806397, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 24769.85713126627}
{"step": 3750, "total_steps": 57221, "tokens": 1966080000, "source": "stack_v3", "source_tokens": {"stack_v3": 1474822144, "fineweb_edu": 491257856}, "loss": 0.9132250919938087, "ce_loss": 0.9112933278083801, "perplexity": 2.48753765545746, "z_loss": 19.317575931549072, "grad_norm": 0.0895039290189743, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177570.037125997, "data_wait_seconds": 1.7702579498291016e-05, "eta_seconds": 23806.583995989058}
{"step": 3760, "total_steps": 57221, "tokens": 1971322880, "source": "stack_v3", "source_tokens": {"stack_v3": 1478492160, "fineweb_edu": 492830720}, "loss": 0.8220160603523254, "ce_loss": 0.8203391134738922, "perplexity": 2.271269925185953, "z_loss": 16.769603967666626, "grad_norm": 0.14548622071743011, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146808.3553059455, "data_wait_seconds": 1.7821788787841797e-05, "eta_seconds": 24440.593749007443}
{"step": 3770, "total_steps": 57221, "tokens": 1976565760, "source": "stack_v3", "source_tokens": {"stack_v3": 1482686464, "fineweb_edu": 493879296}, "loss": 0.8224789649248123, "ce_loss": 0.8206525295972824, "perplexity": 2.271981889365776, "z_loss": 18.264392852783203, "grad_norm": 0.10753725469112396, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157104.478247955, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 24218.585933080176}
{"step": 3780, "total_steps": 57221, "tokens": 1981808640, "source": "stack_v3", "source_tokens": {"stack_v3": 1486356480, "fineweb_edu": 495452160}, "loss": 1.0268298089504242, "ce_loss": 1.0251891314983368, "perplexity": 2.7876226379576687, "z_loss": 16.406922817230225, "grad_norm": 0.22489605844020844, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1115822.611123137, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 25109.897469991352}
{"step": 3790, "total_steps": 57221, "tokens": 1987051520, "source": "stack_v3", "source_tokens": {"stack_v3": 1490550784, "fineweb_edu": 496500736}, "loss": 1.0121254473924637, "ce_loss": 1.0100559294223785, "perplexity": 2.745754579190106, "z_loss": 20.69514274597168, "grad_norm": 0.0866147130727768, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141474.573106261, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 24541.018380960682}
{"step": 3800, "total_steps": 57221, "tokens": 1992294400, "source": "stack_v3", "source_tokens": {"stack_v3": 1494220800, "fineweb_edu": 498073600}, "loss": 0.8532337546348572, "ce_loss": 0.8513689190149307, "perplexity": 2.3428518321685186, "z_loss": 18.648319244384766, "grad_norm": 0.1374330073595047, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1178278.9259203672, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 23770.013180981623}
{"step": 3810, "total_steps": 57221, "tokens": 1997537280, "source": "stack_v3", "source_tokens": {"stack_v3": 1498415104, "fineweb_edu": 499122176}, "loss": 0.7523963004350662, "ce_loss": 0.7510072439908981, "perplexity": 2.119133426409869, "z_loss": 13.890571594238281, "grad_norm": 0.0749279037117958, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145886.096453929, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 24437.387630984187}
{"step": 3820, "total_steps": 57221, "tokens": 2002780160, "source": "stack_v3", "source_tokens": {"stack_v3": 1502085120, "fineweb_edu": 500695040}, "loss": 0.9735277444124222, "ce_loss": 0.9719822630286217, "perplexity": 2.6431787452792705, "z_loss": 15.454700469970703, "grad_norm": 0.31004154682159424, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142633.7098798994, "data_wait_seconds": 2.1517276763916016e-05, "eta_seconds": 24502.35766538233}
{"step": 3830, "total_steps": 57221, "tokens": 2008023040, "source": "stack_v3", "source_tokens": {"stack_v3": 1506279424, "fineweb_edu": 501743616}, "loss": 0.8757292106747627, "ce_loss": 0.8742102980613708, "perplexity": 2.3969816453043435, "z_loss": 15.189040660858154, "grad_norm": 0.09322500228881836, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147012.0940698537, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 24404.2561579959}
{"step": 3840, "total_steps": 57221, "tokens": 2013265920, "source": "stack_v3", "source_tokens": {"stack_v3": 1509949440, "fineweb_edu": 503316480}, "loss": 0.7320207208395004, "ce_loss": 0.7301779836416245, "perplexity": 2.075449970946676, "z_loss": 18.427409410476685, "grad_norm": 0.21189568936824799, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1163934.2230897972, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 24044.94474413339}
{"step": 3850, "total_steps": 57221, "tokens": 2018508800, "source": "stack_v3", "source_tokens": {"stack_v3": 1514143744, "fineweb_edu": 504365056}, "loss": 0.9210464507341385, "ce_loss": 0.9192275702953339, "perplexity": 2.507352887889878, "z_loss": 18.188806533813477, "grad_norm": 0.11850567162036896, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 648078.3194372252, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 43176.09517364879}
{"step": 3860, "total_steps": 57221, "tokens": 2023751680, "source": "stack_v3", "source_tokens": {"stack_v3": 1517813760, "fineweb_edu": 505937920}, "loss": 0.876274399459362, "ce_loss": 0.8744234964251518, "perplexity": 2.3974927323486606, "z_loss": 18.50920295715332, "grad_norm": 0.17029061913490295, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177078.4491483667, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 23767.53082196112}
{"step": 3870, "total_steps": 57221, "tokens": 2028994560, "source": "stack_v3", "source_tokens": {"stack_v3": 1522008064, "fineweb_edu": 506986496}, "loss": 0.7336724400520325, "ce_loss": 0.7317766100168228, "perplexity": 2.078770493440893, "z_loss": 18.958290576934814, "grad_norm": 0.0820111483335495, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144511.353496055, "data_wait_seconds": 2.1457672119140625e-05, "eta_seconds": 24439.255543039413}
{"step": 3880, "total_steps": 57221, "tokens": 2034237440, "source": "stack_v3", "source_tokens": {"stack_v3": 1525678080, "fineweb_edu": 508559360}, "loss": 0.7300461679697037, "ce_loss": 0.7280756458640099, "perplexity": 2.071091257416659, "z_loss": 19.705254077911377, "grad_norm": 0.12567825615406036, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1171060.80827477, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 23880.709150534818}
{"step": 3890, "total_steps": 57221, "tokens": 2039480320, "source": "stack_v3", "source_tokens": {"stack_v3": 1529872384, "fineweb_edu": 509607936}, "loss": 0.7224853187799454, "ce_loss": 0.7208926975727081, "perplexity": 2.0562680170260745, "z_loss": 15.926335573196411, "grad_norm": 0.13788090646266937, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150341.3587767947, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 24306.280450293092}
{"step": 3900, "total_steps": 57221, "tokens": 2044723200, "source": "stack_v3", "source_tokens": {"stack_v3": 1533542400, "fineweb_edu": 511180800}, "loss": 0.7951672449707985, "ce_loss": 0.7934389635920525, "perplexity": 2.2109868704947027, "z_loss": 17.282840728759766, "grad_norm": 0.14402936398983002, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1117508.5441248587, "data_wait_seconds": 2.181529998779297e-05, "eta_seconds": 25015.7163871103}
{"step": 3910, "total_steps": 57221, "tokens": 2049966080, "source": "stack_v3", "source_tokens": {"stack_v3": 1537736704, "fineweb_edu": 512229376}, "loss": 0.9727720618247986, "ce_loss": 0.9709581136703491, "perplexity": 2.640473121181575, "z_loss": 18.139647483825684, "grad_norm": 0.12632261216640472, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1161459.4619293564, "data_wait_seconds": 1.8835067749023438e-05, "eta_seconds": 24064.579811998647}
{"step": 3920, "total_steps": 57221, "tokens": 2055208960, "source": "stack_v3", "source_tokens": {"stack_v3": 1541406720, "fineweb_edu": 513802240}, "loss": 0.8450255393981934, "ce_loss": 0.8435539901256561, "perplexity": 2.3246139683266893, "z_loss": 14.715440273284912, "grad_norm": 0.16206996142864227, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125964.7047602253, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 24818.53198582353}
{"step": 3930, "total_steps": 57221, "tokens": 2060451840, "source": "stack_v3", "source_tokens": {"stack_v3": 1545601024, "fineweb_edu": 514850816}, "loss": 0.9049870371818542, "ce_loss": 0.9032268524169922, "perplexity": 2.467552706588665, "z_loss": 17.60186767578125, "grad_norm": 0.10586276650428772, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131230.4238771768, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 24698.37052670494}
{"step": 3940, "total_steps": 57221, "tokens": 2065694720, "source": "stack_v3", "source_tokens": {"stack_v3": 1549271040, "fineweb_edu": 516423680}, "loss": 0.8464051485061646, "ce_loss": 0.8444733992218971, "perplexity": 2.3267522223689014, "z_loss": 19.317376136779785, "grad_norm": 0.12734249234199524, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128677.9643545814, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 24749.57974037691}
{"step": 3950, "total_steps": 57221, "tokens": 2070937600, "source": "stack_v3", "source_tokens": {"stack_v3": 1553465344, "fineweb_edu": 517472256}, "loss": 0.7751457095146179, "ce_loss": 0.7733945548534393, "perplexity": 2.167110156392132, "z_loss": 17.51157236099243, "grad_norm": 0.09322155267000198, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1106389.665087466, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 25243.423073544403}
{"step": 3960, "total_steps": 57221, "tokens": 2076180480, "source": "stack_v3", "source_tokens": {"stack_v3": 1557135360, "fineweb_edu": 519045120}, "loss": 0.7504853308200836, "ce_loss": 0.7481460645794868, "perplexity": 2.1130788711949453, "z_loss": 23.39259147644043, "grad_norm": 0.12392027676105499, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1082091.0440577148, "data_wait_seconds": 1.901388168334961e-05, "eta_seconds": 25805.425221235495}
{"step": 3970, "total_steps": 57221, "tokens": 2081423360, "source": "stack_v3", "source_tokens": {"stack_v3": 1561329664, "fineweb_edu": 520093696}, "loss": 0.9185420870780945, "ce_loss": 0.9169320464134216, "perplexity": 2.501603800563508, "z_loss": 16.100271224975586, "grad_norm": 0.12496916204690933, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122530.0035222261, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 24871.118413225748}
{"step": 3980, "total_steps": 57221, "tokens": 2086666240, "source": "stack_v3", "source_tokens": {"stack_v3": 1564999680, "fineweb_edu": 521666560}, "loss": 0.8485983610153198, "ce_loss": 0.8466804474592209, "perplexity": 2.331893147808371, "z_loss": 19.1790931224823, "grad_norm": 0.15237979590892792, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1169431.3430620057, "data_wait_seconds": 1.710653305053711e-05, "eta_seconds": 23869.151383365974}
{"step": 3990, "total_steps": 57221, "tokens": 2091909120, "source": "stack_v3", "source_tokens": {"stack_v3": 1569193984, "fineweb_edu": 522715136}, "loss": 0.7855247035622597, "ce_loss": 0.7833860740065575, "perplexity": 2.1888714119916637, "z_loss": 21.38626480102539, "grad_norm": 0.11158423125743866, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1167425.4970103553, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 23905.671883533007}
{"step": 4000, "total_steps": 57221, "tokens": 2097152000, "source": "stack_v3", "source_tokens": {"stack_v3": 1572864000, "fineweb_edu": 524288000}, "loss": 0.7780167534947395, "ce_loss": 0.7765454053878784, "perplexity": 2.1739491652759275, "z_loss": 14.71346378326416, "grad_norm": 0.1742633730173111, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1087799.6694647572, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 25650.72299914318}
{"step": 4010, "total_steps": 57221, "tokens": 2102394880, "source": "stack_v3", "source_tokens": {"stack_v3": 1577058304, "fineweb_edu": 525336576}, "loss": 0.6565853655338287, "ce_loss": 0.6552355289459229, "perplexity": 1.9255959975659906, "z_loss": 13.49824571609497, "grad_norm": 0.09577792137861252, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1064254.7309842703, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 26213.277994262753}
{"step": 4020, "total_steps": 57221, "tokens": 2107637760, "source": "stack_v3", "source_tokens": {"stack_v3": 1580728320, "fineweb_edu": 526909440}, "loss": 0.5625006929039955, "ce_loss": 0.5606223940849304, "perplexity": 1.7517624476581464, "z_loss": 18.782953023910522, "grad_norm": 0.10654211789369583, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1112446.6006247643, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 25072.989772574514}
{"step": 4030, "total_steps": 57221, "tokens": 2112880640, "source": "stack_v3", "source_tokens": {"stack_v3": 1584922624, "fineweb_edu": 527958016}, "loss": 0.6869839057326317, "ce_loss": 0.6855238378047943, "perplexity": 1.9848112824477253, "z_loss": 14.60074496269226, "grad_norm": 0.19161340594291687, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143467.9062538869, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 24388.195949776098}
{"step": 4040, "total_steps": 57221, "tokens": 2118123520, "source": "stack_v3", "source_tokens": {"stack_v3": 1588592640, "fineweb_edu": 529530880}, "loss": 0.8449935168027878, "ce_loss": 0.8433403894305229, "perplexity": 2.3241174821939174, "z_loss": 16.53148651123047, "grad_norm": 0.1806737333536148, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128721.0086542575, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 24702.186161346268}
{"step": 4050, "total_steps": 57221, "tokens": 2123366400, "source": "stack_v3", "source_tokens": {"stack_v3": 1592786944, "fineweb_edu": 530579456}, "loss": 0.6746168434619904, "ce_loss": 0.672961950302124, "perplexity": 1.9600342553363488, "z_loss": 16.548983693122864, "grad_norm": 0.25824570655822754, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1079912.5646188732, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 25813.78762811073}
{"step": 4060, "total_steps": 57221, "tokens": 2128609280, "source": "stack_v3", "source_tokens": {"stack_v3": 1596456960, "fineweb_edu": 532152320}, "loss": 0.7997083365917206, "ce_loss": 0.7976285517215729, "perplexity": 2.220269426306891, "z_loss": 20.797640681266785, "grad_norm": 0.14754605293273926, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1185947.7707178947, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 23501.36440083571}
{"step": 4070, "total_steps": 57221, "tokens": 2133852160, "source": "stack_v3", "source_tokens": {"stack_v3": 1600651264, "fineweb_edu": 533200896}, "loss": 0.6084949374198914, "ce_loss": 0.6067207455635071, "perplexity": 1.834406040692722, "z_loss": 17.74193775653839, "grad_norm": 0.10920675098896027, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1172809.650767753, "data_wait_seconds": 1.722574234008789e-05, "eta_seconds": 23760.16246264521}
{"step": 4080, "total_steps": 57221, "tokens": 2139095040, "source": "stack_v3", "source_tokens": {"stack_v3": 1604321280, "fineweb_edu": 534773760}, "loss": 0.5663950555026531, "ce_loss": 0.5650105103850365, "perplexity": 1.7594662752926449, "z_loss": 13.845332622528076, "grad_norm": 0.20178009569644928, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138758.424432785, "data_wait_seconds": 1.722574234008789e-05, "eta_seconds": 24466.03630956891}
{"step": 4090, "total_steps": 57221, "tokens": 2144337920, "source": "stack_v3", "source_tokens": {"stack_v3": 1608515584, "fineweb_edu": 535822336}, "loss": 0.6382383108139038, "ce_loss": 0.6363710463047028, "perplexity": 1.889611110585399, "z_loss": 18.672659158706665, "grad_norm": 0.09518809616565704, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1093222.5915126754, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 25480.320564411813}
{"step": 4100, "total_steps": 57221, "tokens": 2149580800, "source": "stack_v3", "source_tokens": {"stack_v3": 1612185600, "fineweb_edu": 537395200}, "loss": 0.7637072131037712, "ce_loss": 0.7621463686227798, "perplexity": 2.1428706784196416, "z_loss": 15.608404397964478, "grad_norm": 0.23841796815395355, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1175894.0606866144, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 23684.462853514124}
{"step": 4110, "total_steps": 57221, "tokens": 2154823680, "source": "stack_v3", "source_tokens": {"stack_v3": 1616379904, "fineweb_edu": 538443776}, "loss": 0.5491243712604046, "ce_loss": 0.5470883995294571, "perplexity": 1.7282138171944779, "z_loss": 20.359652757644653, "grad_norm": 0.12370456010103226, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154820.6058737158, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 24112.12285126559}
{"step": 4120, "total_steps": 57221, "tokens": 2160066560, "source": "stack_v3", "source_tokens": {"stack_v3": 1620049920, "fineweb_edu": 540016640}, "loss": 0.6780313551425934, "ce_loss": 0.6760711967945099, "perplexity": 1.9661379690276424, "z_loss": 19.601585149765015, "grad_norm": 0.1165560707449913, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132740.7475154786, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 24577.497985362774}
{"step": 4130, "total_steps": 57221, "tokens": 2165309440, "source": "stack_v3", "source_tokens": {"stack_v3": 1624244224, "fineweb_edu": 541065216}, "loss": 0.6435269936919212, "ce_loss": 0.6418817564845085, "perplexity": 1.9000529543309395, "z_loss": 16.452430725097656, "grad_norm": 0.07360315322875977, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1083586.6425050918, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 25687.554154091747}
{"step": 4140, "total_steps": 57221, "tokens": 2170552320, "source": "stack_v3", "source_tokens": {"stack_v3": 1627914240, "fineweb_edu": 542638080}, "loss": 0.7754528969526291, "ce_loss": 0.773712232708931, "perplexity": 2.167798708662342, "z_loss": 17.40654730796814, "grad_norm": 0.12951123714447021, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128507.9048612723, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 24660.39232877246}
{"step": 4150, "total_steps": 57221, "tokens": 2175795200, "source": "stack_v3", "source_tokens": {"stack_v3": 1632108544, "fineweb_edu": 543686656}, "loss": 0.7711269780993462, "ce_loss": 0.7691628038883209, "perplexity": 2.1579588625511334, "z_loss": 19.641750812530518, "grad_norm": 0.08868343383073807, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134898.7349971354, "data_wait_seconds": 1.913309097290039e-05, "eta_seconds": 24516.90529029467}
{"step": 4160, "total_steps": 57221, "tokens": 2181038080, "source": "stack_v3", "source_tokens": {"stack_v3": 1635778560, "fineweb_edu": 545259520}, "loss": 0.7510886117815971, "ce_loss": 0.7495826631784439, "perplexity": 2.1161166988878337, "z_loss": 15.059488534927368, "grad_norm": 0.13232193887233734, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125960.7843895457, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 24706.86573252408}
{"step": 4170, "total_steps": 57221, "tokens": 2186280960, "source": "stack_v3", "source_tokens": {"stack_v3": 1639972864, "fineweb_edu": 546308096}, "loss": 0.7315227463841438, "ce_loss": 0.7302289679646492, "perplexity": 2.0755557890559255, "z_loss": 12.937690258026123, "grad_norm": 0.12572020292282104, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132021.007056653, "data_wait_seconds": 1.5854835510253906e-05, "eta_seconds": 24569.967223769054}
{"step": 4180, "total_steps": 57221, "tokens": 2191523840, "source": "stack_v3", "source_tokens": {"stack_v3": 1643642880, "fineweb_edu": 547880960}, "loss": 0.6712082549929619, "ce_loss": 0.6701671779155731, "perplexity": 1.9545640532682178, "z_loss": 10.410820960998535, "grad_norm": 0.1255134642124176, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126374.418648916, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 24688.483420420904}
{"step": 4190, "total_steps": 57221, "tokens": 2196766720, "source": "stack_v3", "source_tokens": {"stack_v3": 1647837184, "fineweb_edu": 548929536}, "loss": 0.7331531941890717, "ce_loss": 0.7317027598619461, "perplexity": 2.078616981586506, "z_loss": 14.504240989685059, "grad_norm": 0.1589779555797577, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1097783.168177704, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 25326.707573912576}
{"step": 4200, "total_steps": 57221, "tokens": 2202009600, "source": "stack_v3", "source_tokens": {"stack_v3": 1651507200, "fineweb_edu": 550502400}, "loss": 0.46456649526953697, "ce_loss": 0.46319906041026115, "perplexity": 1.5891496477139448, "z_loss": 13.674397945404053, "grad_norm": 0.13206847012043, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122589.2564147462, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 24762.387704278808}
{"step": 4210, "total_steps": 57221, "tokens": 2207252480, "source": "stack_v3", "source_tokens": {"stack_v3": 1655701504, "fineweb_edu": 551550976}, "loss": 0.8791923075914383, "ce_loss": 0.8773824572563171, "perplexity": 2.4045973253592456, "z_loss": 18.098622798919678, "grad_norm": 0.10710892826318741, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1085718.6043405277, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 25598.48141948483}
{"step": 4220, "total_steps": 57221, "tokens": 2212495360, "source": "stack_v3", "source_tokens": {"stack_v3": 1659371520, "fineweb_edu": 553123840}, "loss": 0.7802731618285179, "ce_loss": 0.778668038547039, "perplexity": 2.1785685627692004, "z_loss": 16.051307201385498, "grad_norm": 0.13180842995643616, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1170340.7680951983, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 23743.088677689895}
{"step": 4230, "total_steps": 57221, "tokens": 2217738240, "source": "stack_v3", "source_tokens": {"stack_v3": 1663565824, "fineweb_edu": 554172416}, "loss": 0.7706540375947952, "ce_loss": 0.7691585868597031, "perplexity": 2.1579497623960417, "z_loss": 14.954548597335815, "grad_norm": 0.10116975009441376, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125537.0265018428, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24683.560918779345}
{"step": 4240, "total_steps": 57221, "tokens": 2222981120, "source": "stack_v3", "source_tokens": {"stack_v3": 1667235840, "fineweb_edu": 555745280}, "loss": 0.7348476126790047, "ce_loss": 0.7334755137562752, "perplexity": 2.0823051260557865, "z_loss": 13.721089839935303, "grad_norm": 0.14201237261295319, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120818.2138994974, "data_wait_seconds": 1.7762184143066406e-05, "eta_seconds": 24782.804682803573}
{"step": 4250, "total_steps": 57221, "tokens": 2228224000, "source": "stack_v3", "source_tokens": {"stack_v3": 1671430144, "fineweb_edu": 556793856}, "loss": 0.7606100216507912, "ce_loss": 0.7592553496360779, "perplexity": 2.1366845450254686, "z_loss": 13.546635150909424, "grad_norm": 0.08113894611597061, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145841.0762601017, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 24237.022546480875}
{"step": 4260, "total_steps": 57221, "tokens": 2233466880, "source": "stack_v3", "source_tokens": {"stack_v3": 1675100160, "fineweb_edu": 558366720}, "loss": 0.9463847726583481, "ce_loss": 0.9449193328619003, "perplexity": 2.572605845466875, "z_loss": 14.65430498123169, "grad_norm": 0.14466923475265503, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1114089.5451791463, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24923.071255942108}
{"step": 4270, "total_steps": 57221, "tokens": 2238709760, "source": "stack_v3", "source_tokens": {"stack_v3": 1679294464, "fineweb_edu": 559415296}, "loss": 0.5727811977267265, "ce_loss": 0.5714964419603348, "perplexity": 1.7709151413088158, "z_loss": 12.847562074661255, "grad_norm": 0.08951475471258163, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145915.0569363683, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 24226.307239753427}
{"step": 4280, "total_steps": 57221, "tokens": 2243952640, "source": "stack_v3", "source_tokens": {"stack_v3": 1682964480, "fineweb_edu": 560988160}, "loss": 0.69862474873662, "ce_loss": 0.6971078291535378, "perplexity": 2.007937004654874, "z_loss": 15.1693434715271, "grad_norm": 0.15945251286029816, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1051783.3305019822, "data_wait_seconds": 1.710653305053711e-05, "eta_seconds": 26389.510610282192}
{"step": 4290, "total_steps": 57221, "tokens": 2249195520, "source": "stack_v3", "source_tokens": {"stack_v3": 1687158784, "fineweb_edu": 562036736}, "loss": 0.9186647981405258, "ce_loss": 0.9169858768582344, "perplexity": 2.5017384666333724, "z_loss": 16.788898229599, "grad_norm": 0.12506800889968872, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1182125.1533961843, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 23475.35233496502}
{"step": 4300, "total_steps": 57221, "tokens": 2254438400, "source": "stack_v3", "source_tokens": {"stack_v3": 1690828800, "fineweb_edu": 563609600}, "loss": 0.7815499007701874, "ce_loss": 0.7803419679403305, "perplexity": 2.182218386643206, "z_loss": 12.079261541366577, "grad_norm": 0.13204897940158844, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098641.6134822834, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 25254.42442695843}
{"step": 4310, "total_steps": 57221, "tokens": 2259681280, "source": "stack_v3", "source_tokens": {"stack_v3": 1695023104, "fineweb_edu": 564658176}, "loss": 0.6315243579447269, "ce_loss": 0.6304007433354855, "perplexity": 1.878363169978405, "z_loss": 11.23595666885376, "grad_norm": 0.10287762433290482, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1076193.5273795614, "data_wait_seconds": 2.0623207092285156e-05, "eta_seconds": 25776.32927002013}
{"step": 4320, "total_steps": 57221, "tokens": 2264924160, "source": "stack_v3", "source_tokens": {"stack_v3": 1698693120, "fineweb_edu": 566231040}, "loss": 0.7087505236268044, "ce_loss": 0.7076024189591408, "perplexity": 2.02912044119082, "z_loss": 11.481035470962524, "grad_norm": 0.11161330342292786, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136412.298813224, "data_wait_seconds": 1.5854835510253906e-05, "eta_seconds": 24405.821609783914}
{"step": 4330, "total_steps": 57221, "tokens": 2270167040, "source": "stack_v3", "source_tokens": {"stack_v3": 1702887424, "fineweb_edu": 567279616}, "loss": 0.7156483381986618, "ce_loss": 0.7142536267638206, "perplexity": 2.042661525268146, "z_loss": 13.94726848602295, "grad_norm": 0.1611790657043457, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1111199.3334855342, "data_wait_seconds": 1.7762184143066406e-05, "eta_seconds": 24954.868244043086}
{"step": 4340, "total_steps": 57221, "tokens": 2275409920, "source": "stack_v3", "source_tokens": {"stack_v3": 1706557440, "fineweb_edu": 568852480}, "loss": 0.6184715032577515, "ce_loss": 0.6172153279185295, "perplexity": 1.8537587375757947, "z_loss": 12.561721563339233, "grad_norm": 0.1473243683576584, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144683.8873844936, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 24220.30255300296}
{"step": 4350, "total_steps": 57221, "tokens": 2280652800, "source": "stack_v3", "source_tokens": {"stack_v3": 1710751744, "fineweb_edu": 569901056}, "loss": 0.3786870911717415, "ce_loss": 0.3772733733057976, "perplexity": 1.4583029159735352, "z_loss": 14.137194871902466, "grad_norm": 0.06819117069244385, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134657.7069826736, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 24429.699837594526}
{"step": 4360, "total_steps": 57221, "tokens": 2285895680, "source": "stack_v3", "source_tokens": {"stack_v3": 1714421760, "fineweb_edu": 571473920}, "loss": 0.555890366435051, "ce_loss": 0.5546540468931198, "perplexity": 1.7413384591072134, "z_loss": 12.36315655708313, "grad_norm": 0.1798674762248993, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132093.2138664147, "data_wait_seconds": 1.710653305053711e-05, "eta_seconds": 24480.408486284083}
{"step": 4370, "total_steps": 57221, "tokens": 2291138560, "source": "stack_v3", "source_tokens": {"stack_v3": 1718616064, "fineweb_edu": 572522496}, "loss": 0.789341002702713, "ce_loss": 0.7881641834974289, "perplexity": 2.199355105431653, "z_loss": 11.76841688156128, "grad_norm": 0.09641041606664658, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149913.1851144554, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 24096.48119413643}
{"step": 4380, "total_steps": 57221, "tokens": 2296381440, "source": "stack_v3", "source_tokens": {"stack_v3": 1722286080, "fineweb_edu": 574095360}, "loss": 0.8034622073173523, "ce_loss": 0.801520474255085, "perplexity": 2.2289273800316574, "z_loss": 19.417240858078003, "grad_norm": 0.23906050622463226, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 806968.4550787173, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 34330.48514554091}
{"step": 4390, "total_steps": 57221, "tokens": 2301624320, "source": "stack_v3", "source_tokens": {"stack_v3": 1726480384, "fineweb_edu": 575143936}, "loss": 1.0182648748159409, "ce_loss": 1.0165798962116241, "perplexity": 2.7637263507535264, "z_loss": 16.849742889404297, "grad_norm": 0.18269018828868866, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119160.6188449997, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 24749.240827098954}
{"step": 4400, "total_steps": 57221, "tokens": 2306867200, "source": "stack_v3", "source_tokens": {"stack_v3": 1730150400, "fineweb_edu": 576716800}, "loss": 0.9353692978620529, "ce_loss": 0.9339354038238525, "perplexity": 2.544503146981283, "z_loss": 14.338826179504395, "grad_norm": 0.13839642703533173, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1169882.704699777, "data_wait_seconds": 1.6570091247558594e-05, "eta_seconds": 23671.717419830387}
{"step": 4410, "total_steps": 57221, "tokens": 2312110080, "source": "stack_v3", "source_tokens": {"stack_v3": 1734344704, "fineweb_edu": 577765376}, "loss": 0.9853596985340118, "ce_loss": 0.9835482090711594, "perplexity": 2.673927082060851, "z_loss": 18.1147403717041, "grad_norm": 0.350162535905838, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165406.7558068864, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 23758.134043791335}
{"step": 4420, "total_steps": 57221, "tokens": 2317352960, "source": "stack_v3", "source_tokens": {"stack_v3": 1738014720, "fineweb_edu": 579338240}, "loss": 0.6195440590381622, "ce_loss": 0.6181948632001877, "perplexity": 1.8555754492842802, "z_loss": 13.49195146560669, "grad_norm": 0.16535960137844086, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099394.8170766856, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 25179.895893641515}
{"step": 4430, "total_steps": 57221, "tokens": 2322595840, "source": "stack_v3", "source_tokens": {"stack_v3": 1742209024, "fineweb_edu": 580386816}, "loss": 0.6831443943083286, "ce_loss": 0.6814979240298271, "perplexity": 1.9768366666937482, "z_loss": 16.46474552154541, "grad_norm": 0.08845751732587814, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146137.7134017837, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 24148.41064591822}
{"step": 4440, "total_steps": 57221, "tokens": 2327838720, "source": "stack_v3", "source_tokens": {"stack_v3": 1745879040, "fineweb_edu": 581959680}, "loss": 0.9625131674110889, "ce_loss": 0.959996409714222, "perplexity": 2.6116870967032453, "z_loss": 25.167356491088867, "grad_norm": 0.24423222243785858, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1113554.2709428656, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 24850.303215638967}
{"step": 4450, "total_steps": 57221, "tokens": 2333081600, "source": "stack_v3", "source_tokens": {"stack_v3": 1750073344, "fineweb_edu": 583008256}, "loss": 0.8395762294530869, "ce_loss": 0.8378285318613052, "perplexity": 2.3113425168950736, "z_loss": 17.47713327407837, "grad_norm": 0.1143319234251976, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118227.0585874608, "data_wait_seconds": 1.6629695892333984e-05, "eta_seconds": 24741.77152800141}
{"step": 4460, "total_steps": 57221, "tokens": 2338324480, "source": "stack_v3", "source_tokens": {"stack_v3": 1753743360, "fineweb_edu": 584581120}, "loss": 0.7561277076601982, "ce_loss": 0.7543612346053123, "perplexity": 2.12625291270377, "z_loss": 17.66503071784973, "grad_norm": 0.419294148683548, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 733639.2930087317, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 37704.7355336661}
{"step": 4470, "total_steps": 57221, "tokens": 2343567360, "source": "stack_v3", "source_tokens": {"stack_v3": 1757937664, "fineweb_edu": 585629696}, "loss": 0.7947107776999474, "ce_loss": 0.7932675182819366, "perplexity": 2.2106078396574866, "z_loss": 14.432497501373291, "grad_norm": 0.16840443015098572, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1171591.5732992468, "data_wait_seconds": 2.2530555725097656e-05, "eta_seconds": 23605.86510717077}
{"step": 4480, "total_steps": 57221, "tokens": 2348810240, "source": "stack_v3", "source_tokens": {"stack_v3": 1761607680, "fineweb_edu": 587202560}, "loss": 0.6674269773066044, "ce_loss": 0.665858369320631, "perplexity": 1.9461603288978693, "z_loss": 15.685956001281738, "grad_norm": 0.11480255424976349, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1085742.191022345, "data_wait_seconds": 1.6868114471435547e-05, "eta_seconds": 25467.546521300217}
{"step": 4490, "total_steps": 57221, "tokens": 2354053120, "source": "stack_v3", "source_tokens": {"stack_v3": 1765801984, "fineweb_edu": 588251136}, "loss": 0.9418883845210075, "ce_loss": 0.9398625046014786, "perplexity": 2.5596294568610465, "z_loss": 20.258747100830078, "grad_norm": 0.1572365164756775, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135469.9161754267, "data_wait_seconds": 2.1278858184814453e-05, "eta_seconds": 24347.581988890655}
{"step": 4500, "total_steps": 57221, "tokens": 2359296000, "source": "stack_v3", "source_tokens": {"stack_v3": 1769472000, "fineweb_edu": 589824000}, "loss": 0.8317889794707298, "ce_loss": 0.830252893269062, "perplexity": 2.2938987784781983, "z_loss": 15.360677480697632, "grad_norm": 0.13334967195987701, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1071745.4887020017, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 25790.3618829092}
{"step": 4510, "total_steps": 57221, "tokens": 2364538880, "source": "stack_v3", "source_tokens": {"stack_v3": 1773666304, "fineweb_edu": 590872576}, "loss": 0.8671513870358467, "ce_loss": 0.8655235655605793, "perplexity": 2.376249882946112, "z_loss": 16.27828621864319, "grad_norm": 0.11778134852647781, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1167323.802086828, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 23674.20339634642}
{"step": 4520, "total_steps": 57221, "tokens": 2369781760, "source": "stack_v3", "source_tokens": {"stack_v3": 1777336320, "fineweb_edu": 592445440}, "loss": 0.6831317394971848, "ce_loss": 0.6816939562559128, "perplexity": 1.9772242283721762, "z_loss": 14.377786636352539, "grad_norm": 0.11557912826538086, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134355.6881672167, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 24357.631850590235}
{"step": 4530, "total_steps": 57221, "tokens": 2375024640, "source": "stack_v3", "source_tokens": {"stack_v3": 1781530624, "fineweb_edu": 593494016}, "loss": 0.6249525286257267, "ce_loss": 0.6234219558537006, "perplexity": 1.865300107786962, "z_loss": 15.305689573287964, "grad_norm": 0.09136087447404861, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1109071.882859846, "data_wait_seconds": 1.6748905181884766e-05, "eta_seconds": 24908.191963866582}
{"step": 4540, "total_steps": 57221, "tokens": 2380267520, "source": "stack_v3", "source_tokens": {"stack_v3": 1785200640, "fineweb_edu": 595066880}, "loss": 0.7123871296644211, "ce_loss": 0.7109460681676865, "perplexity": 2.0359164635738622, "z_loss": 14.41046929359436, "grad_norm": 0.15949949622154236, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151168.2902567643, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 23992.784299018094}
{"step": 4550, "total_steps": 57221, "tokens": 2385510400, "source": "stack_v3", "source_tokens": {"stack_v3": 1789394944, "fineweb_edu": 596115456}, "loss": 0.7810185328125954, "ce_loss": 0.7795757949352264, "perplexity": 2.180547070164661, "z_loss": 14.42749810218811, "grad_norm": 0.0826144814491272, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1109520.0534178559, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 24888.680033257693}
{"step": 4560, "total_steps": 57221, "tokens": 2390753280, "source": "stack_v3", "source_tokens": {"stack_v3": 1793064960, "fineweb_edu": 597688320}, "loss": 0.7954997159540653, "ce_loss": 0.7941411547362804, "perplexity": 2.212539951111081, "z_loss": 13.585631132125854, "grad_norm": 0.14802324771881104, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1063312.4834186174, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 25965.317957365187}
{"step": 4570, "total_steps": 57221, "tokens": 2395996160, "source": "stack_v3", "source_tokens": {"stack_v3": 1797259264, "fineweb_edu": 598736896}, "loss": 0.8659449778497219, "ce_loss": 0.8643633499741554, "perplexity": 2.3734945195115644, "z_loss": 15.816282272338867, "grad_norm": 0.09540065377950668, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128682.9464339104, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24456.827249153746}
{"step": 4580, "total_steps": 57221, "tokens": 2401239040, "source": "stack_v3", "source_tokens": {"stack_v3": 1800929280, "fineweb_edu": 600309760}, "loss": 0.9124757349491119, "ce_loss": 0.9110578373074532, "perplexity": 2.4869519329376533, "z_loss": 14.179169654846191, "grad_norm": 0.1448642611503601, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1124774.1105373749, "data_wait_seconds": 1.913309097290039e-05, "eta_seconds": 24537.158796102045}
{"step": 4590, "total_steps": 57221, "tokens": 2406481920, "source": "stack_v3", "source_tokens": {"stack_v3": 1805123584, "fineweb_edu": 601358336}, "loss": 0.8450513929128647, "ce_loss": 0.8437759056687355, "perplexity": 2.3251298935417153, "z_loss": 12.754894971847534, "grad_norm": 0.12032832950353622, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165699.645661801, "data_wait_seconds": 2.09808349609375e-05, "eta_seconds": 23671.2074012293}
{"step": 4600, "total_steps": 57221, "tokens": 2411724800, "source": "stack_v3", "source_tokens": {"stack_v3": 1808793600, "fineweb_edu": 602931200}, "loss": 0.8786275610327721, "ce_loss": 0.8773752972483635, "perplexity": 2.4045801084849074, "z_loss": 12.522868156433105, "grad_norm": 0.12288149446249008, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1091625.3984102646, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 25272.65785513679}
{"step": 4610, "total_steps": 57221, "tokens": 2416967680, "source": "stack_v3", "source_tokens": {"stack_v3": 1812987904, "fineweb_edu": 603979776}, "loss": 0.6428974717855453, "ce_loss": 0.6417767107486725, "perplexity": 1.8998533723530258, "z_loss": 11.207703590393066, "grad_norm": 0.07478369027376175, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1142090.6517217627, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 24151.35110196122}
{"step": 4620, "total_steps": 57221, "tokens": 2422210560, "source": "stack_v3", "source_tokens": {"stack_v3": 1816657920, "fineweb_edu": 605552640}, "loss": 0.7585010454058647, "ce_loss": 0.756767988204956, "perplexity": 2.131376442619051, "z_loss": 17.330517053604126, "grad_norm": 0.12302881479263306, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1175341.5499218504, "data_wait_seconds": 2.1517276763916016e-05, "eta_seconds": 23463.63866897556}
{"step": 4630, "total_steps": 57221, "tokens": 2427453440, "source": "stack_v3", "source_tokens": {"stack_v3": 1820852224, "fineweb_edu": 606601216}, "loss": 0.6250779330730438, "ce_loss": 0.6236629970371723, "perplexity": 1.8657497761245898, "z_loss": 14.149467945098877, "grad_norm": 0.0644783303141594, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1078969.5050094519, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 25554.51885524648}
{"step": 4640, "total_steps": 57221, "tokens": 2432696320, "source": "stack_v3", "source_tokens": {"stack_v3": 1824522240, "fineweb_edu": 608174080}, "loss": 0.6008099839091301, "ce_loss": 0.5996857210993767, "perplexity": 1.8215462368741462, "z_loss": 11.242557764053345, "grad_norm": 0.10746541619300842, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1104287.893774152, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 24963.873855197802}
{"step": 4650, "total_steps": 57221, "tokens": 2437939200, "source": "stack_v3", "source_tokens": {"stack_v3": 1828716544, "fineweb_edu": 609222656}, "loss": 0.5903203040361404, "ce_loss": 0.589069589972496, "perplexity": 1.8023107470674131, "z_loss": 12.507161140441895, "grad_norm": 0.0836980938911438, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148276.2803677744, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 24002.98714798177}
{"step": 4660, "total_steps": 57221, "tokens": 2443182080, "source": "stack_v3", "source_tokens": {"stack_v3": 1832386560, "fineweb_edu": 610795520}, "loss": 0.7281812354922295, "ce_loss": 0.7270704060792923, "perplexity": 2.069010360162507, "z_loss": 11.108270525932312, "grad_norm": 0.17032285034656525, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147837.9589140993, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 24007.58548364253}
{"step": 4670, "total_steps": 57221, "tokens": 2448424960, "source": "stack_v3", "source_tokens": {"stack_v3": 1836580864, "fineweb_edu": 611844096}, "loss": 0.7605595588684082, "ce_loss": 0.7594614326953888, "perplexity": 2.1371249248891457, "z_loss": 10.981274604797363, "grad_norm": 0.1065901443362236, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1105233.0323529297, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 24928.29496902158}
{"step": 4680, "total_steps": 57221, "tokens": 2453667840, "source": "stack_v3", "source_tokens": {"stack_v3": 1840250880, "fineweb_edu": 613416960}, "loss": 0.7786467969417572, "ce_loss": 0.7774498760700226, "perplexity": 2.1759163280468123, "z_loss": 11.969294786453247, "grad_norm": 0.12862098217010498, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1103835.6278893496, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 24955.103336056927}
{"step": 4690, "total_steps": 57221, "tokens": 2458910720, "source": "stack_v3", "source_tokens": {"stack_v3": 1844445184, "fineweb_edu": 614465536}, "loss": 0.7089734971523285, "ce_loss": 0.7077273055911064, "perplexity": 2.029373867032994, "z_loss": 12.461858987808228, "grad_norm": 0.08989740163087845, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118018.069071534, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 24633.849883009218}
{"step": 4700, "total_steps": 57221, "tokens": 2464153600, "source": "stack_v3", "source_tokens": {"stack_v3": 1848115200, "fineweb_edu": 616038400}, "loss": 0.7760106325149536, "ce_loss": 0.7747795283794403, "perplexity": 2.1701136259694955, "z_loss": 12.311031818389893, "grad_norm": 0.1560804694890976, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165602.8231312097, "data_wait_seconds": 2.09808349609375e-05, "eta_seconds": 23623.695699388627}
{"step": 4710, "total_steps": 57221, "tokens": 2469396480, "source": "stack_v3", "source_tokens": {"stack_v3": 1852309504, "fineweb_edu": 617086976}, "loss": 0.7040072679519653, "ce_loss": 0.7029635608196259, "perplexity": 2.019729437922535, "z_loss": 10.436920881271362, "grad_norm": 0.09336913377046585, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116839.4476482347, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 24650.457662443863}
{"step": 4720, "total_steps": 57221, "tokens": 2474639360, "source": "stack_v3", "source_tokens": {"stack_v3": 1855979520, "fineweb_edu": 618659840}, "loss": 0.7634049654006958, "ce_loss": 0.7620646134018898, "perplexity": 2.1426954947151775, "z_loss": 13.40357494354248, "grad_norm": 0.11290699243545532, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1074694.933574146, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 25612.25495728175}
{"step": 4730, "total_steps": 57221, "tokens": 2479882240, "source": "stack_v3", "source_tokens": {"stack_v3": 1860173824, "fineweb_edu": 619708416}, "loss": 0.8221134468913078, "ce_loss": 0.8203199505805969, "perplexity": 2.271226401499753, "z_loss": 17.935147285461426, "grad_norm": 0.07670753449201584, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1167979.769865272, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23562.15276157949}
{"step": 4740, "total_steps": 57221, "tokens": 2485125120, "source": "stack_v3", "source_tokens": {"stack_v3": 1863843840, "fineweb_edu": 621281280}, "loss": 0.517122283577919, "ce_loss": 0.5157613828778267, "perplexity": 1.6749132667269828, "z_loss": 13.609097719192505, "grad_norm": 0.16718220710754395, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129070.8146095597, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 24369.485530909617}
{"step": 4750, "total_steps": 57221, "tokens": 2490368000, "source": "stack_v3", "source_tokens": {"stack_v3": 1868038144, "fineweb_edu": 622329856}, "loss": 0.4578155614435673, "ce_loss": 0.45666690543293953, "perplexity": 1.5788029059836104, "z_loss": 11.48667025566101, "grad_norm": 0.08758941292762756, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1084537.7940418916, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 25365.30506463605}
{"step": 4760, "total_steps": 57221, "tokens": 2495610880, "source": "stack_v3", "source_tokens": {"stack_v3": 1871708160, "fineweb_edu": 623902720}, "loss": 0.9583464339375496, "ce_loss": 0.9564942568540573, "perplexity": 2.602556566855499, "z_loss": 18.52162754535675, "grad_norm": 0.28367897868156433, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151984.0400045933, "data_wait_seconds": 2.181529998779297e-05, "eta_seconds": 23875.66855517402}
{"step": 4770, "total_steps": 57221, "tokens": 2500853760, "source": "stack_v3", "source_tokens": {"stack_v3": 1875902464, "fineweb_edu": 624951296}, "loss": 0.7523134350776672, "ce_loss": 0.7511111535131931, "perplexity": 2.1193536359926246, "z_loss": 12.022879600524902, "grad_norm": 0.0997065007686615, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119870.6522599556, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 24555.645051064897}
{"step": 4780, "total_steps": 57221, "tokens": 2506096640, "source": "stack_v3", "source_tokens": {"stack_v3": 1879572480, "fineweb_edu": 626524160}, "loss": 0.5652063116431236, "ce_loss": 0.564065620303154, "perplexity": 1.757804558253323, "z_loss": 11.406948328018188, "grad_norm": 0.14835992455482483, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168751.6920820966, "data_wait_seconds": 1.806020736694336e-05, "eta_seconds": 23524.161330641946}
{"step": 4790, "total_steps": 57221, "tokens": 2511339520, "source": "stack_v3", "source_tokens": {"stack_v3": 1883766784, "fineweb_edu": 627572736}, "loss": 0.6682515814900398, "ce_loss": 0.6671805307269096, "perplexity": 1.9487351687770342, "z_loss": 10.710510969161987, "grad_norm": 0.09254365414381027, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1079192.3248849844, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 25471.512209771896}
{"step": 4800, "total_steps": 57221, "tokens": 2516582400, "source": "stack_v3", "source_tokens": {"stack_v3": 1887436800, "fineweb_edu": 629145600}, "loss": 0.958255372941494, "ce_loss": 0.9568903893232346, "perplexity": 2.603587728239269, "z_loss": 13.649862051010132, "grad_norm": 0.12444893270730972, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155194.667707224, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 23791.156909118872}
{"step": 4810, "total_steps": 57221, "tokens": 2521825280, "source": "stack_v3", "source_tokens": {"stack_v3": 1891631104, "fineweb_edu": 630194176}, "loss": 0.6274788789451122, "ce_loss": 0.6261300444602966, "perplexity": 1.8703583517518456, "z_loss": 13.488268613815308, "grad_norm": 0.10074761509895325, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134095.1213394166, "data_wait_seconds": 1.9848346710205078e-05, "eta_seconds": 24229.162265989697}
{"step": 4820, "total_steps": 57221, "tokens": 2527068160, "source": "stack_v3", "source_tokens": {"stack_v3": 1895301120, "fineweb_edu": 631767040}, "loss": 0.5913099758327007, "ce_loss": 0.5900640673935413, "perplexity": 1.8041039959360425, "z_loss": 12.459004640579224, "grad_norm": 0.12133530527353287, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1062555.4190246144, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 25855.528425255325}
{"step": 4830, "total_steps": 57221, "tokens": 2532311040, "source": "stack_v3", "source_tokens": {"stack_v3": 1899495424, "fineweb_edu": 632815616}, "loss": 0.4653483033180237, "ce_loss": 0.46425843983888626, "perplexity": 1.5908340522139464, "z_loss": 10.898692846298218, "grad_norm": 0.12428738921880722, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1113007.0609919394, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 24678.809257077053}
{"step": 4840, "total_steps": 57221, "tokens": 2537553920, "source": "stack_v3", "source_tokens": {"stack_v3": 1903165440, "fineweb_edu": 634388480}, "loss": 0.6325631402432919, "ce_loss": 0.6310419030487537, "perplexity": 1.8795678869366015, "z_loss": 15.212342977523804, "grad_norm": 0.12020254880189896, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1169195.131474044, "data_wait_seconds": 2.1159648895263672e-05, "eta_seconds": 23488.334274345772}
{"step": 4850, "total_steps": 57221, "tokens": 2542796800, "source": "stack_v3", "source_tokens": {"stack_v3": 1907359744, "fineweb_edu": 635437056}, "loss": 0.6541290171444416, "ce_loss": 0.652548037469387, "perplexity": 1.920427922424809, "z_loss": 15.809833288192749, "grad_norm": 0.07240792363882065, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1091498.3381515436, "data_wait_seconds": 1.7344951629638672e-05, "eta_seconds": 25155.51535011851}
{"step": 4860, "total_steps": 57221, "tokens": 2548039680, "source": "stack_v3", "source_tokens": {"stack_v3": 1911029760, "fineweb_edu": 637009920}, "loss": 0.5900088418275118, "ce_loss": 0.5888109728693962, "perplexity": 1.8018446989497254, "z_loss": 11.978779792785645, "grad_norm": 0.15579009056091309, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1016928.0687778314, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 26994.987318023806}
{"step": 4870, "total_steps": 57221, "tokens": 2553282560, "source": "stack_v3", "source_tokens": {"stack_v3": 1915224064, "fineweb_edu": 638058496}, "loss": 0.5321127464994788, "ce_loss": 0.5309963282197714, "perplexity": 1.7006258463407269, "z_loss": 11.164157629013062, "grad_norm": 0.08466222137212753, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130913.8832081223, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 24269.50260981894}
{"step": 4880, "total_steps": 57221, "tokens": 2558525440, "source": "stack_v3", "source_tokens": {"stack_v3": 1918894080, "fineweb_edu": 639631360}, "loss": 0.5681528095155954, "ce_loss": 0.5669262940064073, "perplexity": 1.7628402628483122, "z_loss": 12.265147805213928, "grad_norm": 0.1063646525144577, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1056497.595953854, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 25974.005681692623}
{"step": 4890, "total_steps": 57221, "tokens": 2563768320, "source": "stack_v3", "source_tokens": {"stack_v3": 1923088384, "fineweb_edu": 640679936}, "loss": 0.5943410359323025, "ce_loss": 0.5930989496409893, "perplexity": 1.8095875558963823, "z_loss": 12.420818090438843, "grad_norm": 0.09363241493701935, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1184612.9804477359, "data_wait_seconds": 2.1278858184814453e-05, "eta_seconds": 23160.502318343846}
{"step": 4900, "total_steps": 57221, "tokens": 2569011200, "source": "stack_v3", "source_tokens": {"stack_v3": 1926758400, "fineweb_edu": 642252800}, "loss": 0.47186712734401226, "ce_loss": 0.47047740034759045, "perplexity": 1.6007582133585747, "z_loss": 13.897212982177734, "grad_norm": 0.08436203002929688, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1100581.8886336018, "data_wait_seconds": 2.1278858184814453e-05, "eta_seconds": 24924.077965753382}
{"step": 4910, "total_steps": 57221, "tokens": 2574254080, "source": "stack_v3", "source_tokens": {"stack_v3": 1930952704, "fineweb_edu": 643301376}, "loss": 0.6238445900380611, "ce_loss": 0.6222497299313545, "perplexity": 1.863114835714179, "z_loss": 15.948542356491089, "grad_norm": 0.10608029365539551, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1107861.580508291, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 24755.570914749987}
{"step": 4920, "total_steps": 57221, "tokens": 2579496960, "source": "stack_v3", "source_tokens": {"stack_v3": 1934622720, "fineweb_edu": 644874240}, "loss": 0.8669128492474556, "ce_loss": 0.865664642304182, "perplexity": 2.3765851401895284, "z_loss": 12.482181906700134, "grad_norm": 0.15234346687793732, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1057368.0562881215, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 25932.789322442142}
{"step": 4930, "total_steps": 57221, "tokens": 2584739840, "source": "stack_v3", "source_tokens": {"stack_v3": 1938817024, "fineweb_edu": 645922816}, "loss": 1.0573269687592983, "ce_loss": 1.0559814311563969, "perplexity": 2.8747951834038346, "z_loss": 13.455557107925415, "grad_norm": 0.1046554297208786, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164595.328879249, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23540.589147290448}
{"step": 4940, "total_steps": 57221, "tokens": 2589982720, "source": "stack_v3", "source_tokens": {"stack_v3": 1942487040, "fineweb_edu": 647495680}, "loss": 0.7043081372976303, "ce_loss": 0.7031824663281441, "perplexity": 2.020171616218079, "z_loss": 11.25670576095581, "grad_norm": 0.11794247478246689, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118185.8343940054, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 24512.935539784143}
{"step": 4950, "total_steps": 57221, "tokens": 2595225600, "source": "stack_v3", "source_tokens": {"stack_v3": 1946681344, "fineweb_edu": 648544256}, "loss": 0.925178412348032, "ce_loss": 0.9237380176782608, "perplexity": 2.5186877145142463, "z_loss": 14.404025554656982, "grad_norm": 0.10064597427845001, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131300.5510450632, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 24224.13245948148}
{"step": 4960, "total_steps": 57221, "tokens": 2600468480, "source": "stack_v3", "source_tokens": {"stack_v3": 1950351360, "fineweb_edu": 650117120}, "loss": 0.9293825663626194, "ce_loss": 0.9281728379428387, "perplexity": 2.529882446795691, "z_loss": 12.097251892089844, "grad_norm": 0.13586457073688507, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1104506.5376914018, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 24807.034259181004}
{"step": 4970, "total_steps": 57221, "tokens": 2605711360, "source": "stack_v3", "source_tokens": {"stack_v3": 1954545664, "fineweb_edu": 651165696}, "loss": 0.705771666020155, "ce_loss": 0.7045275643467903, "perplexity": 2.022890773412879, "z_loss": 12.441194772720337, "grad_norm": 0.0836407020688057, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165986.5621612994, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 23494.514884649543}
{"step": 4980, "total_steps": 57221, "tokens": 2610954240, "source": "stack_v3", "source_tokens": {"stack_v3": 1958215680, "fineweb_edu": 652738560}, "loss": 0.7764575108885765, "ce_loss": 0.7753259539604187, "perplexity": 2.1712997556046294, "z_loss": 11.315825700759888, "grad_norm": 0.12992608547210693, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1117826.8291319367, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 24502.04722789605}
{"step": 4990, "total_steps": 57221, "tokens": 2616197120, "source": "stack_v3", "source_tokens": {"stack_v3": 1962409984, "fineweb_edu": 653787136}, "loss": 0.730642169713974, "ce_loss": 0.72968465462327, "perplexity": 2.0744263437631223, "z_loss": 9.575174927711487, "grad_norm": 0.08416102826595306, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120906.4247505346, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 24430.052567585608}
{"step": 5000, "total_steps": 57221, "tokens": 2621440000, "source": "stack_v3", "source_tokens": {"stack_v3": 1966080000, "fineweb_edu": 655360000}, "loss": 0.6485469862818718, "ce_loss": 0.647292822599411, "perplexity": 1.910362133158682, "z_loss": 12.541712045669556, "grad_norm": 0.19752466678619385, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1081478.0887931825, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 25315.871198603418}
{"step": 5010, "total_steps": 57221, "tokens": 2626682880, "source": "stack_v3", "source_tokens": {"stack_v3": 1970274304, "fineweb_edu": 656408576}, "loss": 0.6338509768247604, "ce_loss": 0.6326761990785599, "perplexity": 1.8826421687309745, "z_loss": 11.747763633728027, "grad_norm": 0.12835881114006042, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125595.041676776, "data_wait_seconds": 1.6987323760986328e-05, "eta_seconds": 24318.97450367454}
{"step": 5020, "total_steps": 57221, "tokens": 2631925760, "source": "stack_v3", "source_tokens": {"stack_v3": 1973944320, "fineweb_edu": 657981440}, "loss": 0.6720270588994026, "ce_loss": 0.6709608063101768, "perplexity": 1.9561158665000349, "z_loss": 10.662473797798157, "grad_norm": 0.13154447078704834, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1081049.0929985875, "data_wait_seconds": 2.1696090698242188e-05, "eta_seconds": 25316.217752967263}
{"step": 5030, "total_steps": 57221, "tokens": 2637168640, "source": "stack_v3", "source_tokens": {"stack_v3": 1978138624, "fineweb_edu": 659030016}, "loss": 0.6576895713806152, "ce_loss": 0.6565543487668037, "perplexity": 1.9281371870517996, "z_loss": 11.352291464805603, "grad_norm": 0.09513362497091293, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1026981.8557958993, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 26643.92871750798}
{"step": 5040, "total_steps": 57221, "tokens": 2642411520, "source": "stack_v3", "source_tokens": {"stack_v3": 1981808640, "fineweb_edu": 660602880}, "loss": 0.6870748773217201, "ce_loss": 0.6857140921056271, "perplexity": 1.9851889372546387, "z_loss": 13.607746601104736, "grad_norm": 0.1390901356935501, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138519.9984447123, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24029.080312486505}
{"step": 5050, "total_steps": 57221, "tokens": 2647654400, "source": "stack_v3", "source_tokens": {"stack_v3": 1986002944, "fineweb_edu": 661651456}, "loss": 0.8418956696987152, "ce_loss": 0.8405468910932541, "perplexity": 2.3176341237135043, "z_loss": 13.487748622894287, "grad_norm": 0.11777953058481216, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1161587.9323076603, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 23547.37410680624}
{"step": 5060, "total_steps": 57221, "tokens": 2652897280, "source": "stack_v3", "source_tokens": {"stack_v3": 1989672960, "fineweb_edu": 663224320}, "loss": 0.7572447992861271, "ce_loss": 0.7557757683098316, "perplexity": 2.129262697332164, "z_loss": 14.69038438796997, "grad_norm": 0.10021316260099411, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098772.922740042, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 24888.7665085555}
{"step": 5070, "total_steps": 57221, "tokens": 2658140160, "source": "stack_v3", "source_tokens": {"stack_v3": 1993867264, "fineweb_edu": 664272896}, "loss": 0.869181614369154, "ce_loss": 0.8678761124610901, "perplexity": 2.381846703052684, "z_loss": 13.054887533187866, "grad_norm": 0.10131789743900299, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1180118.120874695, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 23168.74841285753}
{"step": 5080, "total_steps": 57221, "tokens": 2663383040, "source": "stack_v3", "source_tokens": {"stack_v3": 1997537280, "fineweb_edu": 665845760}, "loss": 0.8037748038768768, "ce_loss": 0.8023729175329208, "perplexity": 2.230828224259386, "z_loss": 14.018898487091064, "grad_norm": 0.12041185796260834, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1091443.4594351633, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 25046.296923293736}
{"step": 5090, "total_steps": 57221, "tokens": 2668625920, "source": "stack_v3", "source_tokens": {"stack_v3": 2001731584, "fineweb_edu": 666894336}, "loss": 0.6078526079654694, "ce_loss": 0.6065495004877448, "perplexity": 1.8340919345866344, "z_loss": 13.031066417694092, "grad_norm": 0.09257100522518158, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 711359.907963488, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 38421.302316917805}
{"step": 5100, "total_steps": 57221, "tokens": 2673868800, "source": "stack_v3", "source_tokens": {"stack_v3": 2005401600, "fineweb_edu": 668467200}, "loss": 0.6478541130200028, "ce_loss": 0.6464616768062115, "perplexity": 1.90877500336771, "z_loss": 13.924373865127563, "grad_norm": 0.09173675626516342, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1127396.2520864173, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24238.266846664483}
{"step": 5110, "total_steps": 57221, "tokens": 2679111680, "source": "stack_v3", "source_tokens": {"stack_v3": 2009595904, "fineweb_edu": 669515776}, "loss": 0.2986539015546441, "ce_loss": 0.2972523639909923, "perplexity": 1.3461549776276727, "z_loss": 14.015392303466797, "grad_norm": 0.04542207717895508, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098214.4169067498, "data_wait_seconds": 2.2470951080322266e-05, "eta_seconds": 24877.553872359913}
{"step": 5120, "total_steps": 57221, "tokens": 2684354560, "source": "stack_v3", "source_tokens": {"stack_v3": 2013265920, "fineweb_edu": 671088640}, "loss": 0.4983707112260163, "ce_loss": 0.4972486491315067, "perplexity": 1.6441912946345087, "z_loss": 11.220595836639404, "grad_norm": 0.14477576315402985, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1175930.9090154604, "data_wait_seconds": 1.7344951629638672e-05, "eta_seconds": 23228.954380380925}
{"step": 5130, "total_steps": 57221, "tokens": 2689597440, "source": "stack_v3", "source_tokens": {"stack_v3": 2017460224, "fineweb_edu": 672137216}, "loss": 0.5313604800030589, "ce_loss": 0.5302943796850741, "perplexity": 1.6994325133979744, "z_loss": 10.660902976989746, "grad_norm": 0.08448133617639542, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150519.2064839418, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 23737.459058560427}
{"step": 5140, "total_steps": 57221, "tokens": 2694840320, "source": "stack_v3", "source_tokens": {"stack_v3": 2021130240, "fineweb_edu": 673710080}, "loss": 0.5631853349041194, "ce_loss": 0.5622490323148668, "perplexity": 1.7546142502220003, "z_loss": 9.36313271522522, "grad_norm": 0.13462209701538086, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1065308.7194969587, "data_wait_seconds": 1.6987323760986328e-05, "eta_seconds": 25631.217674528714}
{"step": 5150, "total_steps": 57221, "tokens": 2700083200, "source": "stack_v3", "source_tokens": {"stack_v3": 2025324544, "fineweb_edu": 674758656}, "loss": 0.39980614464730024, "ce_loss": 0.3987459179479629, "perplexity": 1.4899549996855401, "z_loss": 10.602319240570068, "grad_norm": 0.0699218213558197, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1114096.3748285177, "data_wait_seconds": 1.621246337890625e-05, "eta_seconds": 24504.089068777394}
{"step": 5160, "total_steps": 57221, "tokens": 2705326080, "source": "stack_v3", "source_tokens": {"stack_v3": 2028994560, "fineweb_edu": 676331520}, "loss": 0.5218974114395678, "ce_loss": 0.5206928751431406, "perplexity": 1.6831934886912263, "z_loss": 12.04529082775116, "grad_norm": 0.09919939935207367, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1186042.3735461088, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 23013.236734866863}
{"step": 5170, "total_steps": 57221, "tokens": 2710568960, "source": "stack_v3", "source_tokens": {"stack_v3": 2033188864, "fineweb_edu": 677380096}, "loss": 0.5494179581291974, "ce_loss": 0.5482920715585351, "perplexity": 1.7302952722698224, "z_loss": 11.258921384811401, "grad_norm": 0.08419867604970932, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 509002.42056029534, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 53613.558477699524}
{"step": 5180, "total_steps": 57221, "tokens": 2715811840, "source": "stack_v3", "source_tokens": {"stack_v3": 2036858880, "fineweb_edu": 678952960}, "loss": 0.4405699949711561, "ce_loss": 0.43944047717377543, "perplexity": 1.5518386863845695, "z_loss": 11.29522979259491, "grad_norm": 0.09014953672885895, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1065356.9240065059, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 25610.3729606336}
{"step": 5190, "total_steps": 57221, "tokens": 2721054720, "source": "stack_v3", "source_tokens": {"stack_v3": 2041053184, "fineweb_edu": 680001536}, "loss": 0.5454140151850879, "ce_loss": 0.544210312422365, "perplexity": 1.72324701815754, "z_loss": 12.036904215812683, "grad_norm": 0.06814665347337723, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1172081.9614826706, "data_wait_seconds": 1.6570091247558594e-05, "eta_seconds": 23273.92296481761}
{"step": 5200, "total_steps": 57221, "tokens": 2726297600, "source": "stack_v3", "source_tokens": {"stack_v3": 2044723200, "fineweb_edu": 681574400}, "loss": 0.49428500793874264, "ce_loss": 0.493347005918622, "perplexity": 1.6377887451829898, "z_loss": 9.37984585762024, "grad_norm": 0.10710567981004715, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1153620.5863079538, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 23641.830532243475}
{"step": 5210, "total_steps": 57221, "tokens": 2731540480, "source": "stack_v3", "source_tokens": {"stack_v3": 2048917504, "fineweb_edu": 682622976}, "loss": 0.3444344806484878, "ce_loss": 0.3436269210651517, "perplexity": 1.4100524764883495, "z_loss": 8.075671792030334, "grad_norm": 0.0728323832154274, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144326.544958236, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 23829.26415553456}
{"step": 5220, "total_steps": 57221, "tokens": 2736783360, "source": "stack_v3", "source_tokens": {"stack_v3": 2052587520, "fineweb_edu": 684195840}, "loss": 0.583652013912797, "ce_loss": 0.5818206979893148, "perplexity": 1.789293229433366, "z_loss": 18.31305766105652, "grad_norm": 0.36775270104408264, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118083.2704184684, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 24383.887462868588}
{"step": 5230, "total_steps": 57221, "tokens": 2742026240, "source": "stack_v3", "source_tokens": {"stack_v3": 2056781824, "fineweb_edu": 685244416}, "loss": 0.6694545252248645, "ce_loss": 0.6684162365272641, "perplexity": 1.9511447205704013, "z_loss": 10.382802844047546, "grad_norm": 0.09854253381490707, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1172846.8690210576, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 23240.86330447509}
{"step": 5240, "total_steps": 57221, "tokens": 2747269120, "source": "stack_v3", "source_tokens": {"stack_v3": 2060451840, "fineweb_edu": 686817280}, "loss": 0.5283742435276508, "ce_loss": 0.5273346873000264, "perplexity": 1.6944101519179127, "z_loss": 10.395429372787476, "grad_norm": 0.09097939729690552, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152209.3028950696, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23652.58708771411}
{"step": 5250, "total_steps": 57221, "tokens": 2752512000, "source": "stack_v3", "source_tokens": {"stack_v3": 2064646144, "fineweb_edu": 687865856}, "loss": 0.4075844425242394, "ce_loss": 0.4067396295722574, "perplexity": 1.5019130010176112, "z_loss": 8.448024272918701, "grad_norm": 0.11166923493146896, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143403.5157034856, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 23830.15936699812}
{"step": 5260, "total_steps": 57221, "tokens": 2757754880, "source": "stack_v3", "source_tokens": {"stack_v3": 2068316160, "fineweb_edu": 689438720}, "loss": 0.6170845259912312, "ce_loss": 0.6159794791601598, "perplexity": 1.8514691872020448, "z_loss": 11.050422191619873, "grad_norm": 0.09341417253017426, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1065891.9048484142, "data_wait_seconds": 2.1517276763916016e-05, "eta_seconds": 25558.168699924834}
{"step": 5270, "total_steps": 57221, "tokens": 2762997760, "source": "stack_v3", "source_tokens": {"stack_v3": 2072510464, "fineweb_edu": 690487296}, "loss": 0.4670321107842028, "ce_loss": 0.46611957624554634, "perplexity": 1.593797568283875, "z_loss": 9.125243425369263, "grad_norm": 0.08655066788196564, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1181773.2696848481, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 23047.57007007231}
{"step": 5280, "total_steps": 57221, "tokens": 2768240640, "source": "stack_v3", "source_tokens": {"stack_v3": 2076180480, "fineweb_edu": 692060160}, "loss": 0.41890619834885, "ce_loss": 0.41789453849196434, "perplexity": 1.5187604951838503, "z_loss": 10.11660647392273, "grad_norm": 0.13316036760807037, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135165.8827111935, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 23989.23344574147}
{"step": 5290, "total_steps": 57221, "tokens": 2773483520, "source": "stack_v3", "source_tokens": {"stack_v3": 2080374784, "fineweb_edu": 693108736}, "loss": 0.5480693944264203, "ce_loss": 0.5472141865175217, "perplexity": 1.7284312176780638, "z_loss": 8.552123069763184, "grad_norm": 0.09412797540426254, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 539943.2232138264, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 50424.77673475281}
{"step": 5300, "total_steps": 57221, "tokens": 2778726400, "source": "stack_v3", "source_tokens": {"stack_v3": 2084044800, "fineweb_edu": 694681600}, "loss": 0.6647762376815081, "ce_loss": 0.6637312741950154, "perplexity": 1.9420250603625566, "z_loss": 10.449635863304138, "grad_norm": 0.13387148082256317, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1156320.2542847022, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 23541.292733680457}
{"step": 5310, "total_steps": 57221, "tokens": 2783969280, "source": "stack_v3", "source_tokens": {"stack_v3": 2088239104, "fineweb_edu": 695730176}, "loss": 0.5417732736095786, "ce_loss": 0.5407776944339275, "perplexity": 1.7173419102325915, "z_loss": 9.955781102180481, "grad_norm": 0.08185529708862305, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168133.8865692902, "data_wait_seconds": 2.5212764739990234e-05, "eta_seconds": 23298.725456831977}
{"step": 5320, "total_steps": 57221, "tokens": 2789212160, "source": "stack_v3", "source_tokens": {"stack_v3": 2091909120, "fineweb_edu": 697303040}, "loss": 0.47571763629093766, "ce_loss": 0.4746894449926913, "perplexity": 1.607514898158896, "z_loss": 10.281986355781555, "grad_norm": 0.09325278550386429, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137596.598623176, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 23919.540435452247}
{"step": 5330, "total_steps": 57221, "tokens": 2794455040, "source": "stack_v3", "source_tokens": {"stack_v3": 2096103424, "fineweb_edu": 698351616}, "loss": 0.38482902431860566, "ce_loss": 0.38387813745066524, "perplexity": 1.4679665406365618, "z_loss": 9.508970081806183, "grad_norm": 0.1137746274471283, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1091018.4852755242, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 24935.91568536032}
{"step": 5340, "total_steps": 57221, "tokens": 2799697920, "source": "stack_v3", "source_tokens": {"stack_v3": 2099773440, "fineweb_edu": 699924480}, "loss": 0.48336016247048974, "ce_loss": 0.4823388899676502, "perplexity": 1.6198586461391467, "z_loss": 10.212794065475464, "grad_norm": 0.09899275004863739, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1173097.0119374562, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 23186.745685317786}
{"step": 5350, "total_steps": 57221, "tokens": 2804940800, "source": "stack_v3", "source_tokens": {"stack_v3": 2103967744, "fineweb_edu": 700973056}, "loss": 0.6841079918667674, "ce_loss": 0.6829529507085681, "perplexity": 1.9797151103816506, "z_loss": 11.550532937049866, "grad_norm": 0.0844305083155632, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120775.3703741173, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 24264.50466244831}
{"step": 5360, "total_steps": 57221, "tokens": 2810183680, "source": "stack_v3", "source_tokens": {"stack_v3": 2107637760, "fineweb_edu": 702545920}, "loss": 0.4600704303011298, "ce_loss": 0.4589235968887806, "perplexity": 1.582369800188091, "z_loss": 11.468336462974548, "grad_norm": 0.10249573737382889, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1072171.2606299366, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 25359.583229292184}
{"step": 5370, "total_steps": 57221, "tokens": 2815426560, "source": "stack_v3", "source_tokens": {"stack_v3": 2111832064, "fineweb_edu": 703594496}, "loss": 0.530764099676162, "ce_loss": 0.5297655928879976, "perplexity": 1.69853411347431, "z_loss": 9.98509657382965, "grad_norm": 0.06690660119056702, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164068.3560558944, "data_wait_seconds": 1.8596649169921875e-05, "eta_seconds": 23353.07312373561}
{"step": 5380, "total_steps": 57221, "tokens": 2820669440, "source": "stack_v3", "source_tokens": {"stack_v3": 2115502080, "fineweb_edu": 705167360}, "loss": 0.6266267425380647, "ce_loss": 0.6255421375390142, "perplexity": 1.8692590782984704, "z_loss": 10.845966935157776, "grad_norm": 0.12438376247882843, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151493.6193852713, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 23603.544216345526}
{"step": 5390, "total_steps": 57221, "tokens": 2825912320, "source": "stack_v3", "source_tokens": {"stack_v3": 2119696384, "fineweb_edu": 706215936}, "loss": 0.45149998739361763, "ce_loss": 0.4505423866212368, "perplexity": 1.5691630477647993, "z_loss": 9.576047778129578, "grad_norm": 0.07503652572631836, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130077.5122976212, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 24046.215754484754}
{"step": 5400, "total_steps": 57221, "tokens": 2831155200, "source": "stack_v3", "source_tokens": {"stack_v3": 2123366400, "fineweb_edu": 707788800}, "loss": 0.6177596468478441, "ce_loss": 0.6166854584589601, "perplexity": 1.852776747621562, "z_loss": 10.74206018447876, "grad_norm": 0.10184169560670853, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154279.5975506855, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 23537.490273284493}
{"step": 5410, "total_steps": 57221, "tokens": 2836398080, "source": "stack_v3", "source_tokens": {"stack_v3": 2127560704, "fineweb_edu": 708837376}, "loss": 0.6855225395411253, "ce_loss": 0.6845623617991805, "perplexity": 1.9829038511456365, "z_loss": 9.601934790611267, "grad_norm": 0.10987647622823715, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130064.0391381388, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 24037.223537098616}
{"step": 5420, "total_steps": 57221, "tokens": 2841640960, "source": "stack_v3", "source_tokens": {"stack_v3": 2131230720, "fineweb_edu": 710410240}, "loss": 0.7798445783555508, "ce_loss": 0.7787521374411881, "perplexity": 2.1787517856804715, "z_loss": 10.92456203699112, "grad_norm": 0.32894089818000793, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 512935.6155403713, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 52946.91617658292}
{"step": 5430, "total_steps": 57221, "tokens": 2846883840, "source": "stack_v3", "source_tokens": {"stack_v3": 2135425024, "fineweb_edu": 711458816}, "loss": 0.6383443782106042, "ce_loss": 0.637217580806464, "perplexity": 1.891211408843698, "z_loss": 11.267853736877441, "grad_norm": 0.08918985724449158, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150297.0107408043, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 23605.308808472942}
{"step": 5440, "total_steps": 57221, "tokens": 2852126720, "source": "stack_v3", "source_tokens": {"stack_v3": 2139095040, "fineweb_edu": 713031680}, "loss": 0.6610905788838863, "ce_loss": 0.6599397226236761, "perplexity": 1.934675713711212, "z_loss": 11.50850224494934, "grad_norm": 0.1094849556684494, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131126.7330782092, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 24000.735272272024}
{"step": 5450, "total_steps": 57221, "tokens": 2857369600, "source": "stack_v3", "source_tokens": {"stack_v3": 2143289344, "fineweb_edu": 714080256}, "loss": 0.8484790627844632, "ce_loss": 0.8472745241597295, "perplexity": 2.3332788827712316, "z_loss": 12.045349597930908, "grad_norm": 0.10349699854850769, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099777.443984211, "data_wait_seconds": 2.2530555725097656e-05, "eta_seconds": 24680.11191579746}
{"step": 5460, "total_steps": 57221, "tokens": 2862612480, "source": "stack_v3", "source_tokens": {"stack_v3": 2146959360, "fineweb_edu": 715653120}, "loss": 0.7833282235078514, "ce_loss": 0.782124403398484, "perplexity": 2.1861115186705464, "z_loss": 12.038156032562256, "grad_norm": 0.16722284257411957, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141481.6241320886, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 23773.82775709033}
{"step": 5470, "total_steps": 57221, "tokens": 2867855360, "source": "stack_v3", "source_tokens": {"stack_v3": 2151153664, "fineweb_edu": 716701696}, "loss": 0.7188950907438993, "ce_loss": 0.7176854042336345, "perplexity": 2.04968352715138, "z_loss": 12.09700322151184, "grad_norm": 0.07904820889234543, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152419.0117422342, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 23543.645465360256}
{"step": 5480, "total_steps": 57221, "tokens": 2873098240, "source": "stack_v3", "source_tokens": {"stack_v3": 2154823680, "fineweb_edu": 718274560}, "loss": 0.8573348862119019, "ce_loss": 0.8557098233141005, "perplexity": 2.353044033439573, "z_loss": 16.25067949295044, "grad_norm": 0.3161003887653351, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130855.725549412, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 23987.94218141376}
{"step": 5490, "total_steps": 57221, "tokens": 2878341120, "source": "stack_v3", "source_tokens": {"stack_v3": 2159017984, "fineweb_edu": 719323136}, "loss": 0.8485826090909541, "ce_loss": 0.8464947859756649, "perplexity": 2.3314602452549633, "z_loss": 20.878390431404114, "grad_norm": 0.12714463472366333, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136203.912887821, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 23870.414960169004}
{"step": 5500, "total_steps": 57221, "tokens": 2883584000, "source": "stack_v3", "source_tokens": {"stack_v3": 2162688000, "fineweb_edu": 720896000}, "loss": 1.1503249984234571, "ce_loss": 1.1482848399318755, "perplexity": 3.1527807460139234, "z_loss": 20.401775240898132, "grad_norm": 0.22431287169456482, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 843369.2531271449, "data_wait_seconds": 2.086162567138672e-05, "eta_seconds": 32152.483505243436}
{"step": 5510, "total_steps": 57221, "tokens": 2888826880, "source": "stack_v3", "source_tokens": {"stack_v3": 2166882304, "fineweb_edu": 721944576}, "loss": 0.9791866668965667, "ce_loss": 0.9771606104914099, "perplexity": 2.6569015432716214, "z_loss": 20.260679483413696, "grad_norm": 0.1448192000389099, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 969753.3878176978, "data_wait_seconds": 1.7642974853515625e-05, "eta_seconds": 27956.77072189469}
{"step": 5520, "total_steps": 57221, "tokens": 2894069760, "source": "stack_v3", "source_tokens": {"stack_v3": 2170552320, "fineweb_edu": 723517440}, "loss": 0.7616815844085068, "ce_loss": 0.7601695053745061, "perplexity": 2.1386387005286602, "z_loss": 15.12091076374054, "grad_norm": 0.20174670219421387, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151739.6823243296, "data_wait_seconds": 2.205371856689453e-05, "eta_seconds": 23534.77149046166}
{"step": 5530, "total_steps": 57221, "tokens": 2899312640, "source": "stack_v3", "source_tokens": {"stack_v3": 2174746624, "fineweb_edu": 724566016}, "loss": 1.0724970335140824, "ce_loss": 1.0709090381860733, "perplexity": 2.9180308961969943, "z_loss": 15.880020380020142, "grad_norm": 0.0966375395655632, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132275.316165873, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 23934.71532327379}
{"step": 5540, "total_steps": 57221, "tokens": 2904555520, "source": "stack_v3", "source_tokens": {"stack_v3": 2178416640, "fineweb_edu": 726138880}, "loss": 0.67949851276353, "ce_loss": 0.6783133461140096, "perplexity": 1.9705512897500683, "z_loss": 11.851632535457611, "grad_norm": 0.15481561422348022, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141543.4280274862, "data_wait_seconds": 2.2292137145996094e-05, "eta_seconds": 23735.798231363995}
{"step": 5550, "total_steps": 57221, "tokens": 2909798400, "source": "stack_v3", "source_tokens": {"stack_v3": 2182610944, "fineweb_edu": 727187456}, "loss": 1.086922500282526, "ce_loss": 1.0856505082920194, "perplexity": 2.9613655851769227, "z_loss": 12.720023393630981, "grad_norm": 0.1009659394621849, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 468183.8392550901, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 57862.31674100971}
{"step": 5560, "total_steps": 57221, "tokens": 2915041280, "source": "stack_v3", "source_tokens": {"stack_v3": 2186280960, "fineweb_edu": 728760320}, "loss": 1.0029120757244527, "ce_loss": 1.0019007790833712, "perplexity": 2.7234535953373253, "z_loss": 10.112901747226715, "grad_norm": 0.19128885865211487, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1161415.6633398454, "data_wait_seconds": 2.1159648895263672e-05, "eta_seconds": 23320.641846789513}
{"step": 5570, "total_steps": 57221, "tokens": 2920284160, "source": "stack_v3", "source_tokens": {"stack_v3": 2190475264, "fineweb_edu": 729808896}, "loss": 0.7986615309491754, "ce_loss": 0.7977447719313204, "perplexity": 2.2205274814806355, "z_loss": 9.16759604215622, "grad_norm": 0.0942934900522232, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144095.0675987066, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23669.113351599783}
{"step": 5580, "total_steps": 57221, "tokens": 2925527040, "source": "stack_v3", "source_tokens": {"stack_v3": 2194145280, "fineweb_edu": 731381760}, "loss": 0.7499061916023493, "ce_loss": 0.7489706380292773, "perplexity": 2.114821978490679, "z_loss": 9.35560816526413, "grad_norm": 0.2113516628742218, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125241.512730241, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 24061.0328126871}
{"step": 5590, "total_steps": 57221, "tokens": 2930769920, "source": "stack_v3", "source_tokens": {"stack_v3": 2198339584, "fineweb_edu": 732430336}, "loss": 0.8858624952845275, "ce_loss": 0.8846769330557436, "perplexity": 2.422201731754763, "z_loss": 11.855650305747986, "grad_norm": 0.13234420120716095, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165733.8803223285, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 23220.762934774863}
{"step": 5600, "total_steps": 57221, "tokens": 2936012800, "source": "stack_v3", "source_tokens": {"stack_v3": 2202009600, "fineweb_edu": 734003200}, "loss": 0.9577788608148694, "ce_loss": 0.9567698054015636, "perplexity": 2.603273796348533, "z_loss": 10.090641975402832, "grad_norm": 0.1699439287185669, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1162388.2907517317, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 23283.086568686413}
{"step": 5610, "total_steps": 57221, "tokens": 2941255680, "source": "stack_v3", "source_tokens": {"stack_v3": 2206203904, "fineweb_edu": 735051776}, "loss": 1.057507253717631, "ce_loss": 1.056417293380946, "perplexity": 2.8760484711381444, "z_loss": 10.899421334266663, "grad_norm": 0.1337215155363083, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 610715.5634846834, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 44306.62314483267}
{"step": 5620, "total_steps": 57221, "tokens": 2946498560, "source": "stack_v3", "source_tokens": {"stack_v3": 2209873920, "fineweb_edu": 736624640}, "loss": 0.9369061700999737, "ce_loss": 0.9358630524948239, "perplexity": 2.549412785599741, "z_loss": 10.431289792060852, "grad_norm": 0.15701885521411896, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119508.6837125123, "data_wait_seconds": 1.6987323760986328e-05, "eta_seconds": 24165.512812535977}
{"step": 5630, "total_steps": 57221, "tokens": 2951741440, "source": "stack_v3", "source_tokens": {"stack_v3": 2214068224, "fineweb_edu": 737673216}, "loss": 0.8548046771902591, "ce_loss": 0.8537333155982196, "perplexity": 2.348397816905656, "z_loss": 10.713699460029602, "grad_norm": 0.1011991798877716, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157986.470580925, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 23358.009136696346}
{"step": 5640, "total_steps": 57221, "tokens": 2956984320, "source": "stack_v3", "source_tokens": {"stack_v3": 2217738240, "fineweb_edu": 739246080}, "loss": 0.7669899719767272, "ce_loss": 0.7660327740013599, "perplexity": 2.15121494665157, "z_loss": 9.571842432022095, "grad_norm": 0.17488764226436615, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151579.1867025273, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 23483.41824189783}
{"step": 5650, "total_steps": 57221, "tokens": 2962227200, "source": "stack_v3", "source_tokens": {"stack_v3": 2221932544, "fineweb_edu": 740294656}, "loss": 0.8867659624665976, "ce_loss": 0.8858653455972672, "perplexity": 2.425082017815928, "z_loss": 9.006088972091675, "grad_norm": 0.11189702898263931, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152265.9946489104, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 23464.8708940146}
{"step": 5660, "total_steps": 57221, "tokens": 2967470080, "source": "stack_v3", "source_tokens": {"stack_v3": 2225602560, "fineweb_edu": 741867520}, "loss": 0.6864821971394122, "ce_loss": 0.6855490989983082, "perplexity": 1.9848614217829068, "z_loss": 9.331063747406006, "grad_norm": 0.13596481084823608, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130777.8036639795, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 23906.137733167736}
{"step": 5670, "total_steps": 57221, "tokens": 2972712960, "source": "stack_v3", "source_tokens": {"stack_v3": 2229796864, "fineweb_edu": 742916096}, "loss": 0.8704473390243948, "ce_loss": 0.8693753313273191, "perplexity": 2.3854202906920206, "z_loss": 10.72007167339325, "grad_norm": 0.13464102149009705, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168345.5220494063, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 23132.957271570802}
{"step": 5680, "total_steps": 57221, "tokens": 2977955840, "source": "stack_v3", "source_tokens": {"stack_v3": 2233466880, "fineweb_edu": 744488960}, "loss": 0.8699070480652153, "ce_loss": 0.8687424412928522, "perplexity": 2.3839110596010618, "z_loss": 11.645962119102478, "grad_norm": 0.23538890480995178, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148133.7726958299, "data_wait_seconds": 2.1278858184814453e-05, "eta_seconds": 23535.62346358993}
{"step": 5690, "total_steps": 57221, "tokens": 2983198720, "source": "stack_v3", "source_tokens": {"stack_v3": 2237661184, "fineweb_edu": 745537536}, "loss": 0.9216346207540482, "ce_loss": 0.9205511421896517, "perplexity": 2.510673746914459, "z_loss": 10.83477258682251, "grad_norm": 0.10376119613647461, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 509703.9408325048, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 53004.88992859889}
{"step": 5700, "total_steps": 57221, "tokens": 2988441600, "source": "stack_v3", "source_tokens": {"stack_v3": 2241331200, "fineweb_edu": 747110400}, "loss": 0.9052284411154687, "ce_loss": 0.9040253288112581, "perplexity": 2.4695237759979562, "z_loss": 12.031242370605469, "grad_norm": 0.1726962924003601, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126224.3171086363, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 23984.172593029216}
{"step": 5710, "total_steps": 57221, "tokens": 2993684480, "source": "stack_v3", "source_tokens": {"stack_v3": 2245525504, "fineweb_edu": 748158976}, "loss": 0.7198339840397239, "ce_loss": 0.7188678220845759, "perplexity": 2.0521085429511126, "z_loss": 9.661539673805237, "grad_norm": 0.09655439853668213, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1102214.8095965355, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 24501.862327440176}
{"step": 5720, "total_steps": 57221, "tokens": 2998927360, "source": "stack_v3", "source_tokens": {"stack_v3": 2249195520, "fineweb_edu": 749731840}, "loss": 0.8673955826088786, "ce_loss": 0.8662794819101691, "perplexity": 2.378046808160351, "z_loss": 11.161155641078949, "grad_norm": 0.19865892827510834, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151038.8607609977, "data_wait_seconds": 2.2232532501220703e-05, "eta_seconds": 23458.002644800814}
{"step": 5730, "total_steps": 57221, "tokens": 3004170240, "source": "stack_v3", "source_tokens": {"stack_v3": 2253389824, "fineweb_edu": 750780416}, "loss": 0.9273830275051296, "ce_loss": 0.926410349085927, "perplexity": 2.525427484238079, "z_loss": 9.726727426052094, "grad_norm": 0.10313763469457626, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1180432.4655190527, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 22869.440267495145}
{"step": 5740, "total_steps": 57221, "tokens": 3009413120, "source": "stack_v3", "source_tokens": {"stack_v3": 2257059840, "fineweb_edu": 752353280}, "loss": 0.8890183875337243, "ce_loss": 0.8879258120432496, "perplexity": 2.430083969349385, "z_loss": 10.925985217094421, "grad_norm": 0.14605052769184113, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122566.1619419586, "data_wait_seconds": 1.704692840576172e-05, "eta_seconds": 24043.649091745498}
{"step": 5750, "total_steps": 57221, "tokens": 3014656000, "source": "stack_v3", "source_tokens": {"stack_v3": 2261254144, "fineweb_edu": 753401856}, "loss": 0.8911951133050025, "ce_loss": 0.8903616690076888, "perplexity": 2.436010521496575, "z_loss": 8.334643542766571, "grad_norm": 0.18474118411540985, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1179658.1321182752, "data_wait_seconds": 2.6047229766845703e-05, "eta_seconds": 22875.563068041807}
{"step": 5760, "total_steps": 57221, "tokens": 3019898880, "source": "stack_v3", "source_tokens": {"stack_v3": 2264924160, "fineweb_edu": 754974720}, "loss": 0.9281932632438838, "ce_loss": 0.9273035135120153, "perplexity": 2.5276841138482937, "z_loss": 8.89747667312622, "grad_norm": 0.1651146113872528, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147842.572342809, "data_wait_seconds": 2.2351741790771484e-05, "eta_seconds": 23505.053541386034}
{"step": 5770, "total_steps": 57221, "tokens": 3025141760, "source": "stack_v3", "source_tokens": {"stack_v3": 2269118464, "fineweb_edu": 756023296}, "loss": 0.7733312945347279, "ce_loss": 0.7725676582194865, "perplexity": 2.165318920983735, "z_loss": 7.636494338512421, "grad_norm": 0.10040491074323654, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1114630.5320441737, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 24200.71715649986}
{"step": 5780, "total_steps": 57221, "tokens": 3030384640, "source": "stack_v3", "source_tokens": {"stack_v3": 2272788480, "fineweb_edu": 757596160}, "loss": 0.9438547343015671, "ce_loss": 0.9429333284497261, "perplexity": 2.567501709003411, "z_loss": 9.213962078094482, "grad_norm": 0.18681669235229492, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150949.4581262688, "data_wait_seconds": 1.6689300537109375e-05, "eta_seconds": 23432.493207743624}
{"step": 5790, "total_steps": 57221, "tokens": 3035627520, "source": "stack_v3", "source_tokens": {"stack_v3": 2276982784, "fineweb_edu": 758644736}, "loss": 0.8169080698862672, "ce_loss": 0.8158629587851465, "perplexity": 2.261126089287021, "z_loss": 10.451235830783844, "grad_norm": 0.08809807151556015, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1156234.2241563746, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 23320.856550215045}
{"step": 5800, "total_steps": 57221, "tokens": 3040870400, "source": "stack_v3", "source_tokens": {"stack_v3": 2280652800, "fineweb_edu": 760217600}, "loss": 0.7610197891481221, "ce_loss": 0.7600979567505419, "perplexity": 2.138485689346416, "z_loss": 9.218409180641174, "grad_norm": 0.1598409116268158, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150438.1299985524, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23433.79352354561}
{"step": 5810, "total_steps": 57221, "tokens": 3046113280, "source": "stack_v3", "source_tokens": {"stack_v3": 2284847104, "fineweb_edu": 761266176}, "loss": 0.8101590271107852, "ce_loss": 0.8091807225719094, "perplexity": 2.246067080611926, "z_loss": 9.78317266702652, "grad_norm": 0.10738483816385269, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1096149.0262157293, "data_wait_seconds": 2.104043960571289e-05, "eta_seconds": 24589.61881584092}
{"step": 5820, "total_steps": 57221, "tokens": 3051356160, "source": "stack_v3", "source_tokens": {"stack_v3": 2288517120, "fineweb_edu": 762839040}, "loss": 0.719315514434129, "ce_loss": 0.7181758806109428, "perplexity": 2.0506890950859056, "z_loss": 11.396344661712646, "grad_norm": 0.13243110477924347, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 501555.8457145688, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 53730.0962001672}
{"step": 5830, "total_steps": 57221, "tokens": 3056599040, "source": "stack_v3", "source_tokens": {"stack_v3": 2292711424, "fineweb_edu": 763887616}, "loss": 0.6376268267631531, "ce_loss": 0.6367684118449688, "perplexity": 1.890362126129216, "z_loss": 8.584132671356201, "grad_norm": 0.13032680749893188, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1186206.6690822088, "data_wait_seconds": 1.6629695892333984e-05, "eta_seconds": 22713.91795566841}
{"step": 5840, "total_steps": 57221, "tokens": 3061841920, "source": "stack_v3", "source_tokens": {"stack_v3": 2296381440, "fineweb_edu": 765460480}, "loss": 0.7053319048136473, "ce_loss": 0.7044553894083947, "perplexity": 2.022744776664643, "z_loss": 8.765069484710693, "grad_norm": 0.1270906925201416, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125584.3254146064, "data_wait_seconds": 2.0623207092285156e-05, "eta_seconds": 23932.598803805653}
{"step": 5850, "total_steps": 57221, "tokens": 3067084800, "source": "stack_v3", "source_tokens": {"stack_v3": 2300575744, "fineweb_edu": 766509056}, "loss": 0.7418923527002335, "ce_loss": 0.7409789003431797, "perplexity": 2.0979882310273275, "z_loss": 9.13433051109314, "grad_norm": 0.11230498552322388, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128309.5861749006, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23870.146571479272}
{"step": 5860, "total_steps": 57221, "tokens": 3072327680, "source": "stack_v3", "source_tokens": {"stack_v3": 2304245760, "fineweb_edu": 768081920}, "loss": 0.7866098750382662, "ce_loss": 0.7857232629321516, "perplexity": 2.1939932009781087, "z_loss": 8.866231322288513, "grad_norm": 0.1315525472164154, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129124.846384677, "data_wait_seconds": 2.3543834686279297e-05, "eta_seconds": 23848.26833473658}
{"step": 5870, "total_steps": 57221, "tokens": 3077570560, "source": "stack_v3", "source_tokens": {"stack_v3": 2308440064, "fineweb_edu": 769130496}, "loss": 0.7900102315470576, "ce_loss": 0.7892106752842665, "perplexity": 2.201657917212475, "z_loss": 7.9954134821891785, "grad_norm": 0.10786797851324081, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149618.6174306793, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 23418.574674938573}
{"step": 5880, "total_steps": 57221, "tokens": 3082813440, "source": "stack_v3", "source_tokens": {"stack_v3": 2312110080, "fineweb_edu": 770703360}, "loss": 0.8622628117445856, "ce_loss": 0.8613508711569011, "perplexity": 2.366355176507595, "z_loss": 9.119495034217834, "grad_norm": 0.1542970836162567, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120897.1116800322, "data_wait_seconds": 2.187490463256836e-05, "eta_seconds": 24013.96727631474}
{"step": 5890, "total_steps": 57221, "tokens": 3088056320, "source": "stack_v3", "source_tokens": {"stack_v3": 2316304384, "fineweb_edu": 771751936}, "loss": 0.7191412441898137, "ce_loss": 0.7182335304096341, "perplexity": 2.050807320307213, "z_loss": 9.077259004116058, "grad_norm": 0.08339102566242218, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126468.93045698, "data_wait_seconds": 1.6391277313232422e-05, "eta_seconds": 23890.53346467577}
{"step": 5900, "total_steps": 57221, "tokens": 3093299200, "source": "stack_v3", "source_tokens": {"stack_v3": 2319974400, "fineweb_edu": 773324800}, "loss": 0.8195681378711015, "ce_loss": 0.8186148232780397, "perplexity": 2.2673569712253876, "z_loss": 9.53329211473465, "grad_norm": 0.13602314889431, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1114640.1367642395, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 24139.361137765227}
{"step": 5910, "total_steps": 57221, "tokens": 3098542080, "source": "stack_v3", "source_tokens": {"stack_v3": 2324168704, "fineweb_edu": 774373376}, "loss": 0.8426773664541543, "ce_loss": 0.8416567747481167, "perplexity": 2.3202078559528356, "z_loss": 10.205996751785278, "grad_norm": 0.10163552314043045, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1164731.711761452, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 23096.699135388244}
{"step": 5920, "total_steps": 57221, "tokens": 3103784960, "source": "stack_v3", "source_tokens": {"stack_v3": 2327838720, "fineweb_edu": 775946240}, "loss": 0.785663434304297, "ce_loss": 0.7848942885175347, "perplexity": 2.1921751903948956, "z_loss": 7.691599905490875, "grad_norm": 0.1777692437171936, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1104572.7247722703, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 24349.881575742504}
{"step": 5930, "total_steps": 57221, "tokens": 3109027840, "source": "stack_v3", "source_tokens": {"stack_v3": 2332033024, "fineweb_edu": 776994816}, "loss": 0.749142961576581, "ce_loss": 0.7482973141595721, "perplexity": 2.1133984976579754, "z_loss": 8.456463634967804, "grad_norm": 0.09892760217189789, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134186.1949657314, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 23709.48639593739}
{"step": 5940, "total_steps": 57221, "tokens": 3114270720, "source": "stack_v3", "source_tokens": {"stack_v3": 2335703040, "fineweb_edu": 778567680}, "loss": 1.176407556515187, "ce_loss": 1.175293382955715, "perplexity": 3.239093099158725, "z_loss": 11.141645908355713, "grad_norm": 0.16074159741401672, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151129.6029723224, "data_wait_seconds": 1.627206802368164e-05, "eta_seconds": 23355.953326696297}
{"step": 5950, "total_steps": 57221, "tokens": 3119513600, "source": "stack_v3", "source_tokens": {"stack_v3": 2339897344, "fineweb_edu": 779616256}, "loss": 0.6737838056869805, "ce_loss": 0.6729286105837673, "perplexity": 1.9599689094356194, "z_loss": 8.552119851112366, "grad_norm": 0.09687028080224991, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148234.309772274, "data_wait_seconds": 1.6987323760986328e-05, "eta_seconds": 23410.27974101482}
{"step": 5960, "total_steps": 57221, "tokens": 3124756480, "source": "stack_v3", "source_tokens": {"stack_v3": 2343567360, "fineweb_edu": 781189120}, "loss": 0.6763554806821048, "ce_loss": 0.6756999446079135, "perplexity": 1.9654081714853395, "z_loss": 6.555344820022583, "grad_norm": 0.16026192903518677, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1105605.2262607976, "data_wait_seconds": 2.0444393157958984e-05, "eta_seconds": 24308.173371152723}
{"step": 5970, "total_steps": 57221, "tokens": 3129999360, "source": "stack_v3", "source_tokens": {"stack_v3": 2347761664, "fineweb_edu": 782237696}, "loss": 1.0245663970708847, "ce_loss": 1.023610532283783, "perplexity": 2.7832255705675957, "z_loss": 9.55886459350586, "grad_norm": 0.13299450278282166, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1092344.6453235676, "data_wait_seconds": 1.901388168334961e-05, "eta_seconds": 24598.464188965503}
{"step": 5980, "total_steps": 57221, "tokens": 3135242240, "source": "stack_v3", "source_tokens": {"stack_v3": 2351431680, "fineweb_edu": 783810560}, "loss": 1.2788327634334564, "ce_loss": 1.2779695391654968, "perplexity": 3.5893442985165933, "z_loss": 8.632392764091492, "grad_norm": 0.31671401858329773, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1114937.9652095374, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 24095.293727800483}
{"step": 5990, "total_steps": 57221, "tokens": 3140485120, "source": "stack_v3", "source_tokens": {"stack_v3": 2355625984, "fineweb_edu": 784859136}, "loss": 0.8173309788107872, "ce_loss": 0.8162349760532379, "perplexity": 2.261967423723334, "z_loss": 10.959952592849731, "grad_norm": 0.11192843317985535, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140352.5563343244, "data_wait_seconds": 2.2113323211669922e-05, "eta_seconds": 23553.693750939798}
{"step": 6000, "total_steps": 57221, "tokens": 3145728000, "source": "stack_v3", "source_tokens": {"stack_v3": 2359296000, "fineweb_edu": 786432000}, "loss": 1.267247587442398, "ce_loss": 1.265780508518219, "perplexity": 3.545859229665804, "z_loss": 14.670911312103271, "grad_norm": 0.37814080715179443, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1109089.0554050447, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 24212.908665114082}
{"step": 6010, "total_steps": 57221, "tokens": 3150970880, "source": "stack_v3", "source_tokens": {"stack_v3": 2363490304, "fineweb_edu": 787480576}, "loss": 0.9653058499097824, "ce_loss": 0.9642462581396103, "perplexity": 2.6228099895620556, "z_loss": 10.59598159790039, "grad_norm": 0.12025915831327438, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1160179.563482703, "data_wait_seconds": 2.0623207092285156e-05, "eta_seconds": 23142.13244663854}
{"step": 6020, "total_steps": 57221, "tokens": 3156213760, "source": "stack_v3", "source_tokens": {"stack_v3": 2367160320, "fineweb_edu": 789053440}, "loss": 1.107394628226757, "ce_loss": 1.1062748730182648, "perplexity": 3.023076051235528, "z_loss": 11.197730302810669, "grad_norm": 0.19376270473003387, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147713.4103554492, "data_wait_seconds": 2.2292137145996094e-05, "eta_seconds": 23388.92793078581}
{"step": 6030, "total_steps": 57221, "tokens": 3161456640, "source": "stack_v3", "source_tokens": {"stack_v3": 2371354624, "fineweb_edu": 790102016}, "loss": 1.0646341517567635, "ce_loss": 1.0633938536047935, "perplexity": 2.8961835516361187, "z_loss": 12.403278112411499, "grad_norm": 0.12052574753761292, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1139368.0908351215, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 23555.6389334444}
{"step": 6040, "total_steps": 57221, "tokens": 3166699520, "source": "stack_v3", "source_tokens": {"stack_v3": 2375024640, "fineweb_edu": 791674880}, "loss": 1.12691081315279, "ce_loss": 1.1257269233465195, "perplexity": 3.0824567444754063, "z_loss": 11.839110851287842, "grad_norm": 0.18704473972320557, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1106585.6981686829, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 24248.73240672378}
{"step": 6050, "total_steps": 57221, "tokens": 3171942400, "source": "stack_v3", "source_tokens": {"stack_v3": 2379218944, "fineweb_edu": 792723456}, "loss": 0.8909813240170479, "ce_loss": 0.8898487240076065, "perplexity": 2.4347613024971357, "z_loss": 11.326014399528503, "grad_norm": 0.1171351820230484, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1169852.6445742233, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 22932.852034338284}
{"step": 6060, "total_steps": 57221, "tokens": 3177185280, "source": "stack_v3", "source_tokens": {"stack_v3": 2382888960, "fineweb_edu": 794296320}, "loss": 0.9440386965870857, "ce_loss": 0.9431684091687202, "perplexity": 2.568105350100594, "z_loss": 8.702735662460327, "grad_norm": 0.3430059850215912, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145990.7790940348, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 23405.78581374348}
{"step": 6070, "total_steps": 57221, "tokens": 3182428160, "source": "stack_v3", "source_tokens": {"stack_v3": 2387083264, "fineweb_edu": 795344896}, "loss": 1.0459659546613693, "ce_loss": 1.0449524223804474, "perplexity": 2.843263244736724, "z_loss": 10.135278940200806, "grad_norm": 0.11377283930778503, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108723.1212116505, "data_wait_seconds": 1.71661376953125e-05, "eta_seconds": 24187.79885341693}
{"step": 6080, "total_steps": 57221, "tokens": 3187671040, "source": "stack_v3", "source_tokens": {"stack_v3": 2390753280, "fineweb_edu": 796917760}, "loss": 1.3175675719976425, "ce_loss": 1.3163122981786728, "perplexity": 3.729642177834793, "z_loss": 12.552589416503906, "grad_norm": 0.34089457988739014, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1114221.976284601, "data_wait_seconds": 1.7702579498291016e-05, "eta_seconds": 24063.72296605236}
{"step": 6090, "total_steps": 57221, "tokens": 3192913920, "source": "stack_v3", "source_tokens": {"stack_v3": 2394947584, "fineweb_edu": 797966336}, "loss": 1.1649069786071777, "ce_loss": 1.163919597864151, "perplexity": 3.2024610676856136, "z_loss": 9.873671293258667, "grad_norm": 0.13809612393379211, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1106968.388232814, "data_wait_seconds": 1.6808509826660156e-05, "eta_seconds": 24216.668122560717}
{"step": 6100, "total_steps": 57221, "tokens": 3198156800, "source": "stack_v3", "source_tokens": {"stack_v3": 2398617600, "fineweb_edu": 799539200}, "loss": 1.167611226439476, "ce_loss": 1.1668711155653, "perplexity": 3.2119271509984104, "z_loss": 7.401067674160004, "grad_norm": 0.33185380697250366, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144801.7002400467, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 23411.777947552033}
{"step": 6110, "total_steps": 57221, "tokens": 3203399680, "source": "stack_v3", "source_tokens": {"stack_v3": 2402811904, "fineweb_edu": 800587776}, "loss": 1.0419772788882256, "ce_loss": 1.0409945920109749, "perplexity": 2.8320323308048954, "z_loss": 9.826768398284912, "grad_norm": 0.12267204374074936, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099089.1379113868, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 24380.73436966352}
{"step": 6120, "total_steps": 57221, "tokens": 3208642560, "source": "stack_v3", "source_tokens": {"stack_v3": 2406481920, "fineweb_edu": 802160640}, "loss": 0.8783183638006449, "ce_loss": 0.8775078393518925, "perplexity": 2.4048988377126443, "z_loss": 8.105155348777771, "grad_norm": 0.20866289734840393, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 879925.2657852381, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 30447.31010887795}
{"step": 6130, "total_steps": 57221, "tokens": 3213885440, "source": "stack_v3", "source_tokens": {"stack_v3": 2410676224, "fineweb_edu": 803209216}, "loss": 1.1033311784267426, "ce_loss": 1.1022585779428482, "perplexity": 3.010958835224024, "z_loss": 10.726165533065796, "grad_norm": 0.19917374849319458, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148192.0424877733, "data_wait_seconds": 1.7821788787841797e-05, "eta_seconds": 23328.94983487506}
{"step": 6140, "total_steps": 57221, "tokens": 3219128320, "source": "stack_v3", "source_tokens": {"stack_v3": 2414346240, "fineweb_edu": 804782080}, "loss": 1.1444186121225357, "ce_loss": 1.1432728990912437, "perplexity": 3.1370187275779386, "z_loss": 11.457001209259033, "grad_norm": 0.29694923758506775, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1115341.0498590723, "data_wait_seconds": 2.1457672119140625e-05, "eta_seconds": 24011.374532824615}
{"step": 6150, "total_steps": 57221, "tokens": 3224371200, "source": "stack_v3", "source_tokens": {"stack_v3": 2418540544, "fineweb_edu": 805830656}, "loss": 0.9116066284477711, "ce_loss": 0.9102780073881149, "perplexity": 2.485013289416749, "z_loss": 13.286061763763428, "grad_norm": 0.10179587453603745, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1105666.65289936, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 24216.7281881813}
{"step": 6160, "total_steps": 57221, "tokens": 3229614080, "source": "stack_v3", "source_tokens": {"stack_v3": 2422210560, "fineweb_edu": 807403520}, "loss": 0.8314450457692146, "ce_loss": 0.8304963856935501, "perplexity": 2.2944573934597985, "z_loss": 9.486449956893921, "grad_norm": 0.1738864928483963, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1110814.4424313495, "data_wait_seconds": 2.0682811737060547e-05, "eta_seconds": 24099.782013461227}
{"step": 6170, "total_steps": 57221, "tokens": 3234856960, "source": "stack_v3", "source_tokens": {"stack_v3": 2426404864, "fineweb_edu": 808452096}, "loss": 0.8672078885138035, "ce_loss": 0.8661851659417152, "perplexity": 2.377822530949238, "z_loss": 10.227174997329712, "grad_norm": 0.10930510610342026, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1086908.005579501, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 24625.030731768114}
{"step": 6180, "total_steps": 57221, "tokens": 3240099840, "source": "stack_v3", "source_tokens": {"stack_v3": 2430074880, "fineweb_edu": 810024960}, "loss": 0.9653786644339561, "ce_loss": 0.9644138962030411, "perplexity": 2.6232497092052975, "z_loss": 9.647740364074707, "grad_norm": 0.146852508187294, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1049618.8861966934, "data_wait_seconds": 1.704692840576172e-05, "eta_seconds": 25494.87295999867}
{"step": 6190, "total_steps": 57221, "tokens": 3245342720, "source": "stack_v3", "source_tokens": {"stack_v3": 2434269184, "fineweb_edu": 811073536}, "loss": 0.8606420792639256, "ce_loss": 0.8595968410372734, "perplexity": 2.362208156316659, "z_loss": 10.452497482299805, "grad_norm": 0.14810103178024292, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1124177.0273055572, "data_wait_seconds": 1.8835067749023438e-05, "eta_seconds": 23799.32753485092}
{"step": 6200, "total_steps": 57221, "tokens": 3250585600, "source": "stack_v3", "source_tokens": {"stack_v3": 2437939200, "fineweb_edu": 812646400}, "loss": 0.8765108287334442, "ce_loss": 0.875408411026001, "perplexity": 2.399855221180074, "z_loss": 11.024177551269531, "grad_norm": 0.18243058025836945, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 299047.8189067156, "data_wait_seconds": 1.8298625946044922e-05, "eta_seconds": 89448.61894593574}
{"step": 6210, "total_steps": 57221, "tokens": 3255828480, "source": "stack_v3", "source_tokens": {"stack_v3": 2442133504, "fineweb_edu": 813694976}, "loss": 0.8170849829912186, "ce_loss": 0.8159414604306221, "perplexity": 2.2613035983729444, "z_loss": 11.435309648513794, "grad_norm": 0.10983084887266159, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108061.5972697733, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 24135.997119561533}
{"step": 6220, "total_steps": 57221, "tokens": 3261071360, "source": "stack_v3", "source_tokens": {"stack_v3": 2445803520, "fineweb_edu": 815267840}, "loss": 0.911351777613163, "ce_loss": 0.9103513658046722, "perplexity": 2.4851955927434446, "z_loss": 10.00446367263794, "grad_norm": 0.18879075348377228, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1153527.9987251002, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 23180.129714712035}
{"step": 6230, "total_steps": 57221, "tokens": 3266314240, "source": "stack_v3", "source_tokens": {"stack_v3": 2449997824, "fineweb_edu": 816316416}, "loss": 0.9848344400525093, "ce_loss": 0.983562782406807, "perplexity": 2.673966050381664, "z_loss": 12.716811656951904, "grad_norm": 0.0943826287984848, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119009.1309731633, "data_wait_seconds": 1.806020736694336e-05, "eta_seconds": 23890.49831680162}
{"step": 6240, "total_steps": 57221, "tokens": 3271557120, "source": "stack_v3", "source_tokens": {"stack_v3": 2453667840, "fineweb_edu": 817889280}, "loss": 1.1121522039175034, "ce_loss": 1.1110830456018448, "perplexity": 3.03764652322449, "z_loss": 10.69138765335083, "grad_norm": 0.423771470785141, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165214.211982811, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 22938.6516274265}
{"step": 6250, "total_steps": 57221, "tokens": 3276800000, "source": "stack_v3", "source_tokens": {"stack_v3": 2457862144, "fineweb_edu": 818937856}, "loss": 1.5261272713541985, "ce_loss": 1.524650163948536, "perplexity": 4.593536303580631, "z_loss": 14.771265745162964, "grad_norm": 0.1558157354593277, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116405.4660298773, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 23936.82296722545}
{"step": 6260, "total_steps": 57221, "tokens": 3282042880, "source": "stack_v3", "source_tokens": {"stack_v3": 2461532160, "fineweb_edu": 820510720}, "loss": 1.083563357591629, "ce_loss": 1.0823878645896912, "perplexity": 2.9517194489098255, "z_loss": 11.75489592552185, "grad_norm": 0.15905886888504028, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144495.8661194448, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 23344.73885920667}
{"step": 6270, "total_steps": 57221, "tokens": 3287285760, "source": "stack_v3", "source_tokens": {"stack_v3": 2465726464, "fineweb_edu": 821559296}, "loss": 1.126253642141819, "ce_loss": 1.1250029653310776, "perplexity": 3.0802259827943215, "z_loss": 12.506641149520874, "grad_norm": 0.14179979264736176, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1107813.5269563345, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 24113.006015905878}
{"step": 6280, "total_steps": 57221, "tokens": 3292528640, "source": "stack_v3", "source_tokens": {"stack_v3": 2469396480, "fineweb_edu": 823132160}, "loss": 0.9555395469069481, "ce_loss": 0.9545152708888054, "perplexity": 2.5974112368830764, "z_loss": 10.242630243301392, "grad_norm": 0.22379688918590546, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1156574.650030471, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23091.86991025298}
{"step": 6290, "total_steps": 57221, "tokens": 3297771520, "source": "stack_v3", "source_tokens": {"stack_v3": 2473590784, "fineweb_edu": 824180736}, "loss": 1.0522768646478653, "ce_loss": 1.05132594704628, "perplexity": 2.8614427253021546, "z_loss": 9.509366273880005, "grad_norm": 0.14416788518428802, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1053784.9482725968, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 25339.352705474943}
{"step": 6300, "total_steps": 57221, "tokens": 3303014400, "source": "stack_v3", "source_tokens": {"stack_v3": 2477260800, "fineweb_edu": 825753600}, "loss": 0.730021558701992, "ce_loss": 0.7291247844696045, "perplexity": 2.073265259425834, "z_loss": 8.967862486839294, "grad_norm": 0.1356240063905716, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1177625.7834834456, "data_wait_seconds": 2.282857894897461e-05, "eta_seconds": 22670.177550825756}
{"step": 6310, "total_steps": 57221, "tokens": 3308257280, "source": "stack_v3", "source_tokens": {"stack_v3": 2481455104, "fineweb_edu": 826802176}, "loss": 0.6193981096148491, "ce_loss": 0.6184337213635445, "perplexity": 1.8560187215655601, "z_loss": 9.643736243247986, "grad_norm": 0.07257098704576492, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1163089.3481987168, "data_wait_seconds": 1.7642974853515625e-05, "eta_seconds": 22949.00452947803}
{"step": 6320, "total_steps": 57221, "tokens": 3313500160, "source": "stack_v3", "source_tokens": {"stack_v3": 2485125120, "fineweb_edu": 828375040}, "loss": 0.9303185790777206, "ce_loss": 0.9291316270828247, "perplexity": 2.5323092338137316, "z_loss": 11.869604229927063, "grad_norm": 0.14307205379009247, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165470.3747093377, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 22897.62178352708}
{"step": 6330, "total_steps": 57221, "tokens": 3318743040, "source": "stack_v3", "source_tokens": {"stack_v3": 2489319424, "fineweb_edu": 829423616}, "loss": 0.8988327831029892, "ce_loss": 0.8979094475507736, "perplexity": 2.4544665528418173, "z_loss": 9.233544826507568, "grad_norm": 0.11533774435520172, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1081630.6506743766, "data_wait_seconds": 1.6450881958007812e-05, "eta_seconds": 24667.622855699152}
{"step": 6340, "total_steps": 57221, "tokens": 3323985920, "source": "stack_v3", "source_tokens": {"stack_v3": 2492989440, "fineweb_edu": 830996480}, "loss": 1.1095218658447266, "ce_loss": 1.1085722297430038, "perplexity": 3.0300291191117505, "z_loss": 9.496183633804321, "grad_norm": 0.20081664621829987, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155572.1296144098, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 23084.681082522497}
{"step": 6350, "total_steps": 57221, "tokens": 3329228800, "source": "stack_v3", "source_tokens": {"stack_v3": 2497183744, "fineweb_edu": 832045056}, "loss": 1.0963119715452194, "ce_loss": 1.0953077226877213, "perplexity": 2.9901026642650206, "z_loss": 10.04240345954895, "grad_norm": 0.13394339382648468, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128727.4974471729, "data_wait_seconds": 2.205371856689453e-05, "eta_seconds": 23629.061275038403}
{"step": 6360, "total_steps": 57221, "tokens": 3334471680, "source": "stack_v3", "source_tokens": {"stack_v3": 2500853760, "fineweb_edu": 833617920}, "loss": 1.1918270587921143, "ce_loss": 1.1907316595315933, "perplexity": 3.2894871117237146, "z_loss": 10.954051733016968, "grad_norm": 0.276754230260849, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1065553.2450220801, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 25025.054772788426}
{"step": 6370, "total_steps": 57221, "tokens": 3339714560, "source": "stack_v3", "source_tokens": {"stack_v3": 2505048064, "fineweb_edu": 834666496}, "loss": 1.0775552242994308, "ce_loss": 1.076221153140068, "perplexity": 2.93357305607876, "z_loss": 13.34045922756195, "grad_norm": 0.14890031516551971, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1180144.087442439, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 22590.703731590183}
{"step": 6380, "total_steps": 57221, "tokens": 3344957440, "source": "stack_v3", "source_tokens": {"stack_v3": 2508718080, "fineweb_edu": 836239360}, "loss": 1.1208883821964264, "ce_loss": 1.119762361049652, "perplexity": 3.0641259610899456, "z_loss": 11.260067462921143, "grad_norm": 0.17003273963928223, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1124648.3046384833, "data_wait_seconds": 1.728534698486328e-05, "eta_seconds": 23700.78045737883}
{"step": 6390, "total_steps": 57221, "tokens": 3350200320, "source": "stack_v3", "source_tokens": {"stack_v3": 2512912384, "fineweb_edu": 837287936}, "loss": 0.9995851740241051, "ce_loss": 0.9983898773789406, "perplexity": 2.7139085830720924, "z_loss": 11.95296573638916, "grad_norm": 0.13333672285079956, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1090917.0559342334, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 24428.804678626824}
{"step": 6400, "total_steps": 57221, "tokens": 3355443200, "source": "stack_v3", "source_tokens": {"stack_v3": 2516582400, "fineweb_edu": 838860800}, "loss": 1.084560178220272, "ce_loss": 1.083413302898407, "perplexity": 2.95474780754143, "z_loss": 11.46869707107544, "grad_norm": 0.177195206284523, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118799.678289731, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 23815.30609727255}
{"step": 6410, "total_steps": 57221, "tokens": 3360686080, "source": "stack_v3", "source_tokens": {"stack_v3": 2520776704, "fineweb_edu": 839909376}, "loss": 1.0453379601240158, "ce_loss": 1.0442233830690384, "perplexity": 2.8411911494698563, "z_loss": 11.14564847946167, "grad_norm": 0.13031364977359772, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152512.2670253061, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 23114.126141804503}
{"step": 6420, "total_steps": 57221, "tokens": 3365928960, "source": "stack_v3", "source_tokens": {"stack_v3": 2524446720, "fineweb_edu": 841482240}, "loss": 1.2234758138656616, "ce_loss": 1.2221155017614365, "perplexity": 3.394360920007135, "z_loss": 13.603052735328674, "grad_norm": 0.13020657002925873, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138366.2890350292, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 23396.749619647628}
{"step": 6430, "total_steps": 57221, "tokens": 3371171840, "source": "stack_v3", "source_tokens": {"stack_v3": 2528641024, "fineweb_edu": 842530816}, "loss": 0.9401044398546219, "ce_loss": 0.9387466758489609, "perplexity": 2.556774941588571, "z_loss": 13.577836632728577, "grad_norm": 0.0993303582072258, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131567.7541036136, "data_wait_seconds": 1.806020736694336e-05, "eta_seconds": 23532.685571350856}
{"step": 6440, "total_steps": 57221, "tokens": 3376414720, "source": "stack_v3", "source_tokens": {"stack_v3": 2532311040, "fineweb_edu": 844103680}, "loss": 1.1318185552954674, "ce_loss": 1.130358211696148, "perplexity": 3.09676559912457, "z_loss": 14.603432774543762, "grad_norm": 0.14282558858394623, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1121581.5487005552, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 23737.53857741831}
{"step": 6450, "total_steps": 57221, "tokens": 3381657600, "source": "stack_v3", "source_tokens": {"stack_v3": 2536505344, "fineweb_edu": 845152256}, "loss": 1.0884162187576294, "ce_loss": 1.0872766375541687, "perplexity": 2.9661850658982045, "z_loss": 11.395642280578613, "grad_norm": 0.15454882383346558, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116834.3426900106, "data_wait_seconds": 1.996755599975586e-05, "eta_seconds": 23833.74273384805}
{"step": 6460, "total_steps": 57221, "tokens": 3386900480, "source": "stack_v3", "source_tokens": {"stack_v3": 2540175360, "fineweb_edu": 846725120}, "loss": 1.1177224665880203, "ce_loss": 1.1166156083345413, "perplexity": 3.054499069070576, "z_loss": 11.068620562553406, "grad_norm": 0.1964098960161209, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165035.1255445846, "data_wait_seconds": 2.181529998779297e-05, "eta_seconds": 22843.173511665547}
{"step": 6470, "total_steps": 57221, "tokens": 3392143360, "source": "stack_v3", "source_tokens": {"stack_v3": 2544369664, "fineweb_edu": 847773696}, "loss": 0.8725349456071854, "ce_loss": 0.8715486004948616, "perplexity": 2.390610088432459, "z_loss": 9.863369703292847, "grad_norm": 0.1096854880452156, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154151.9511060426, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23054.032542683188}
{"step": 6480, "total_steps": 57221, "tokens": 3397386240, "source": "stack_v3", "source_tokens": {"stack_v3": 2548039680, "fineweb_edu": 849346560}, "loss": 0.9332196153700352, "ce_loss": 0.9321133717894554, "perplexity": 2.539871201793543, "z_loss": 11.062458038330078, "grad_norm": 0.15131734311580658, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1056569.779210482, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 25178.28380429233}
{"step": 6490, "total_steps": 57221, "tokens": 3402629120, "source": "stack_v3", "source_tokens": {"stack_v3": 2552233984, "fineweb_edu": 850395136}, "loss": 0.9188944548368454, "ce_loss": 0.9179442673921585, "perplexity": 2.5041372584042008, "z_loss": 9.5019770860672, "grad_norm": 0.1729390025138855, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166169.5899978813, "data_wait_seconds": 1.6689300537109375e-05, "eta_seconds": 22807.463947030483}
{"step": 6500, "total_steps": 57221, "tokens": 3407872000, "source": "stack_v3", "source_tokens": {"stack_v3": 2555904000, "fineweb_edu": 851968000}, "loss": 0.8456611633300781, "ce_loss": 0.8445916250348091, "perplexity": 2.3270273208033836, "z_loss": 10.695383906364441, "grad_norm": 0.16179800033569336, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132510.3753395306, "data_wait_seconds": 1.6510486602783203e-05, "eta_seconds": 23480.692609131802}
{"step": 6510, "total_steps": 57221, "tokens": 3413114880, "source": "stack_v3", "source_tokens": {"stack_v3": 2560098304, "fineweb_edu": 853016576}, "loss": 1.0142782479524612, "ce_loss": 1.0134288966655731, "perplexity": 2.755031556117938, "z_loss": 8.4937082529068, "grad_norm": 0.10756280273199081, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1153553.2319067037, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 23047.81815404794}
{"step": 6520, "total_steps": 57221, "tokens": 3418357760, "source": "stack_v3", "source_tokens": {"stack_v3": 2563768320, "fineweb_edu": 854589440}, "loss": 0.8680904880166054, "ce_loss": 0.8670791052281857, "perplexity": 2.379949110300705, "z_loss": 10.113860368728638, "grad_norm": 0.13698482513427734, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136646.31763155, "data_wait_seconds": 1.817941665649414e-05, "eta_seconds": 23386.027674280103}
{"step": 6530, "total_steps": 57221, "tokens": 3423600640, "source": "stack_v3", "source_tokens": {"stack_v3": 2567962624, "fineweb_edu": 855638016}, "loss": 0.9320640675723553, "ce_loss": 0.931031096726656, "perplexity": 2.53712384949273, "z_loss": 10.329661846160889, "grad_norm": 0.1162361204624176, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1105038.2000741207, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 24050.20872420282}
{"step": 6540, "total_steps": 57221, "tokens": 3428843520, "source": "stack_v3", "source_tokens": {"stack_v3": 2571632640, "fineweb_edu": 857210880}, "loss": 1.004163220524788, "ce_loss": 1.0032674968242645, "perplexity": 2.727178332434917, "z_loss": 8.957298040390015, "grad_norm": 0.18630293011665344, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1167380.5656241432, "data_wait_seconds": 2.1517276763916016e-05, "eta_seconds": 22761.349008576013}
{"step": 6550, "total_steps": 57221, "tokens": 3434086400, "source": "stack_v3", "source_tokens": {"stack_v3": 2575826944, "fineweb_edu": 858259456}, "loss": 1.0109233856201172, "ce_loss": 1.009908102452755, "perplexity": 2.7453487126110865, "z_loss": 10.152616500854492, "grad_norm": 0.10461510717868805, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1127089.5363231387, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 23570.366633572845}
{"step": 6560, "total_steps": 57221, "tokens": 3439329280, "source": "stack_v3", "source_tokens": {"stack_v3": 2579496960, "fineweb_edu": 859832320}, "loss": 0.9752393141388893, "ce_loss": 0.9740001261234283, "perplexity": 2.6485177029535905, "z_loss": 12.391817450523376, "grad_norm": 0.1493563950061798, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1086625.767450116, "data_wait_seconds": 2.3543834686279297e-05, "eta_seconds": 24443.254996913485}
{"step": 6570, "total_steps": 57221, "tokens": 3444572160, "source": "stack_v3", "source_tokens": {"stack_v3": 2583691264, "fineweb_edu": 860880896}, "loss": 1.015420962125063, "ce_loss": 1.0143746957182884, "perplexity": 2.757638494979735, "z_loss": 10.462642431259155, "grad_norm": 0.11931188404560089, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1121703.1791312979, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 23674.20217224117}
{"step": 6580, "total_steps": 57221, "tokens": 3449815040, "source": "stack_v3", "source_tokens": {"stack_v3": 2587361280, "fineweb_edu": 862453760}, "loss": 1.1737884432077408, "ce_loss": 1.1726181954145432, "perplexity": 3.2304394978192876, "z_loss": 11.702591180801392, "grad_norm": 0.13232868909835815, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1180506.7254362043, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 22490.49868834042}
{"step": 6590, "total_steps": 57221, "tokens": 3455057920, "source": "stack_v3", "source_tokens": {"stack_v3": 2591555584, "fineweb_edu": 863502336}, "loss": 1.1463572978973389, "ce_loss": 1.1452963650226593, "perplexity": 3.1433728045605367, "z_loss": 10.609232902526855, "grad_norm": 0.1608102172613144, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1092588.92968805, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 24295.452167521933}
{"step": 6600, "total_steps": 57221, "tokens": 3460300800, "source": "stack_v3", "source_tokens": {"stack_v3": 2595225600, "fineweb_edu": 865075200}, "loss": 0.8737570382654667, "ce_loss": 0.8726029433310032, "perplexity": 2.3931319402677977, "z_loss": 11.541005253791809, "grad_norm": 0.16540180146694183, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1095631.1214699775, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 24223.20677089971}
{"step": 6610, "total_steps": 57221, "tokens": 3465543680, "source": "stack_v3", "source_tokens": {"stack_v3": 2599419904, "fineweb_edu": 866123776}, "loss": 0.9004919826984406, "ce_loss": 0.899534061551094, "perplexity": 2.458457354445194, "z_loss": 9.579387307167053, "grad_norm": 0.14749768376350403, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154446.968029612, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 22984.560620648088}
{"step": 6620, "total_steps": 57221, "tokens": 3470786560, "source": "stack_v3", "source_tokens": {"stack_v3": 2603089920, "fineweb_edu": 867696640}, "loss": 0.9139207080006599, "ce_loss": 0.9128224775195122, "perplexity": 2.491344382730916, "z_loss": 10.982269883155823, "grad_norm": 0.14137090742588043, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136298.8484758667, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 23347.03891989682}
{"step": 6630, "total_steps": 57221, "tokens": 3476029440, "source": "stack_v3", "source_tokens": {"stack_v3": 2607284224, "fineweb_edu": 868745216}, "loss": 0.9665607437491417, "ce_loss": 0.9655686467885971, "perplexity": 2.6262806580012095, "z_loss": 9.92112648487091, "grad_norm": 0.1134827584028244, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1056004.455413841, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 25117.290390224192}
{"step": 6640, "total_steps": 57221, "tokens": 3481272320, "source": "stack_v3", "source_tokens": {"stack_v3": 2610954240, "fineweb_edu": 870318080}, "loss": 1.038110688328743, "ce_loss": 1.0371443033218384, "perplexity": 2.8211491538858287, "z_loss": 9.663697838783264, "grad_norm": 0.16480398178100586, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165585.7093982266, "data_wait_seconds": 1.8596649169921875e-05, "eta_seconds": 22751.417991982074}
{"step": 6650, "total_steps": 57221, "tokens": 3486515200, "source": "stack_v3", "source_tokens": {"stack_v3": 2615148544, "fineweb_edu": 871366656}, "loss": 0.9899633079767227, "ce_loss": 0.9888959638774395, "perplexity": 2.6882648918412175, "z_loss": 10.67349898815155, "grad_norm": 0.11900606751441956, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140097.9756097307, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 23255.44415235054}
{"step": 6660, "total_steps": 57221, "tokens": 3491758080, "source": "stack_v3", "source_tokens": {"stack_v3": 2618818560, "fineweb_edu": 872939520}, "loss": 0.8247198909521103, "ce_loss": 0.8236771076917648, "perplexity": 2.278864078632298, "z_loss": 10.42772114276886, "grad_norm": 0.1616729199886322, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1074906.5860584867, "data_wait_seconds": 1.728534698486328e-05, "eta_seconds": 24660.972649913285}
{"step": 6670, "total_steps": 57221, "tokens": 3497000960, "source": "stack_v3", "source_tokens": {"stack_v3": 2623012864, "fineweb_edu": 873988096}, "loss": 0.7641381248831749, "ce_loss": 0.7630784213542938, "perplexity": 2.1448688779577556, "z_loss": 10.59708547592163, "grad_norm": 0.09169770777225494, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1076986.878017025, "data_wait_seconds": 1.6629695892333984e-05, "eta_seconds": 24608.469778942876}
{"step": 6680, "total_steps": 57221, "tokens": 3502243840, "source": "stack_v3", "source_tokens": {"stack_v3": 2626682880, "fineweb_edu": 875560960}, "loss": 0.7856766395270824, "ce_loss": 0.7846965566277504, "perplexity": 2.1917417703036595, "z_loss": 9.800912022590637, "grad_norm": 0.12715938687324524, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1180179.3658054722, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 22452.312697329093}
{"step": 6690, "total_steps": 57221, "tokens": 3507486720, "source": "stack_v3", "source_tokens": {"stack_v3": 2630877184, "fineweb_edu": 876609536}, "loss": 0.7387640327215195, "ce_loss": 0.7378102205693722, "perplexity": 2.091350899496599, "z_loss": 9.538029074668884, "grad_norm": 0.08986788243055344, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1140018.8930870588, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 23238.6616052133}
{"step": 6700, "total_steps": 57221, "tokens": 3512729600, "source": "stack_v3", "source_tokens": {"stack_v3": 2634547200, "fineweb_edu": 878182400}, "loss": 0.7912672851234674, "ce_loss": 0.7903046887367964, "perplexity": 2.2040678786165055, "z_loss": 9.625968217849731, "grad_norm": 0.12437063455581665, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125834.2517917291, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 23526.793893369613}
{"step": 6710, "total_steps": 57221, "tokens": 3517972480, "source": "stack_v3", "source_tokens": {"stack_v3": 2638741504, "fineweb_edu": 879230976}, "loss": 0.7335194367915392, "ce_loss": 0.7324509844183922, "perplexity": 2.080172835847171, "z_loss": 10.684415221214294, "grad_norm": 0.08600560575723648, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1163978.7663469403, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 22751.29777763202}
{"step": 6720, "total_steps": 57221, "tokens": 3523215360, "source": "stack_v3", "source_tokens": {"stack_v3": 2642411520, "fineweb_edu": 880803840}, "loss": 0.9274523891508579, "ce_loss": 0.9263722291216254, "perplexity": 2.5253312168673996, "z_loss": 10.801813006401062, "grad_norm": 0.16167689859867096, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1139299.0848929451, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 23239.53823107644}
{"step": 6730, "total_steps": 57221, "tokens": 3528458240, "source": "stack_v3", "source_tokens": {"stack_v3": 2646605824, "fineweb_edu": 881852416}, "loss": 0.8060655100271106, "ce_loss": 0.8051799712702632, "perplexity": 2.2370990761551632, "z_loss": 8.85539424419403, "grad_norm": 0.1114741638302803, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148971.399615237, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 23039.339159238152}
{"step": 6740, "total_steps": 57221, "tokens": 3533701120, "source": "stack_v3", "source_tokens": {"stack_v3": 2650275840, "fineweb_edu": 883425280}, "loss": 1.2623946070671082, "ce_loss": 1.2612627148628235, "perplexity": 3.5298759012106182, "z_loss": 11.319268465042114, "grad_norm": 0.45998239517211914, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1090298.9289163356, "data_wait_seconds": 1.913309097290039e-05, "eta_seconds": 24274.350985839497}
{"step": 6750, "total_steps": 57221, "tokens": 3538944000, "source": "stack_v3", "source_tokens": {"stack_v3": 2654470144, "fineweb_edu": 884473856}, "loss": 1.2821149379014969, "ce_loss": 1.2812481820583344, "perplexity": 3.601131789618701, "z_loss": 8.667272806167603, "grad_norm": 0.1592187136411667, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137811.9716794922, "data_wait_seconds": 1.901388168334961e-05, "eta_seconds": 23256.08857932966}
{"step": 6760, "total_steps": 57221, "tokens": 3544186880, "source": "stack_v3", "source_tokens": {"stack_v3": 2658140160, "fineweb_edu": 886046720}, "loss": 0.741459347307682, "ce_loss": 0.7405178919434547, "perplexity": 2.0970212637374472, "z_loss": 9.414515972137451, "grad_norm": 0.14279033243656158, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1169802.610027044, "data_wait_seconds": 1.704692840576172e-05, "eta_seconds": 22615.621552928817}
{"step": 6770, "total_steps": 57221, "tokens": 3549429760, "source": "stack_v3", "source_tokens": {"stack_v3": 2662334464, "fineweb_edu": 887095296}, "loss": 1.095346923917532, "ce_loss": 1.0942331552505493, "perplexity": 2.9868913230185767, "z_loss": 11.137685418128967, "grad_norm": 0.12128254026174545, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1103533.787725502, "data_wait_seconds": 1.9788742065429688e-05, "eta_seconds": 23968.971801504493}
{"step": 6780, "total_steps": 57221, "tokens": 3554672640, "source": "stack_v3", "source_tokens": {"stack_v3": 2666004480, "fineweb_edu": 888668160}, "loss": 0.975667517632246, "ce_loss": 0.9745497517287731, "perplexity": 2.6499737962157575, "z_loss": 11.177737951278687, "grad_norm": 0.17948849499225616, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135509.375677316, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 23289.395866261097}
{"step": 6790, "total_steps": 57221, "tokens": 3559915520, "source": "stack_v3", "source_tokens": {"stack_v3": 2670198784, "fineweb_edu": 889716736}, "loss": 0.8290428295731544, "ce_loss": 0.827909491956234, "perplexity": 2.288529546658773, "z_loss": 11.33341932296753, "grad_norm": 0.11612071096897125, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1178277.852635703, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 22439.600660282187}
{"step": 6800, "total_steps": 57221, "tokens": 3565158400, "source": "stack_v3", "source_tokens": {"stack_v3": 2673868800, "fineweb_edu": 891289600}, "loss": 0.9704644791781902, "ce_loss": 0.9695856608450413, "perplexity": 2.636851682081875, "z_loss": 8.788198232650757, "grad_norm": 0.15529005229473114, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151796.9914856965, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 22950.955589753576}
{"step": 6810, "total_steps": 57221, "tokens": 3570401280, "source": "stack_v3", "source_tokens": {"stack_v3": 2678063104, "fineweb_edu": 892338176}, "loss": 0.9849275834858418, "ce_loss": 0.9838413000106812, "perplexity": 2.6747109007210077, "z_loss": 10.8627188205719, "grad_norm": 0.10417746007442474, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1038899.8226528411, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 25439.987709798384}
{"step": 6820, "total_steps": 57221, "tokens": 3575644160, "source": "stack_v3", "source_tokens": {"stack_v3": 2681733120, "fineweb_edu": 893911040}, "loss": 0.8551756665110588, "ce_loss": 0.8542634844779968, "perplexity": 2.349643194446553, "z_loss": 9.121720671653748, "grad_norm": 0.19468030333518982, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1151493.9811653395, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 22947.88880551327}
{"step": 6830, "total_steps": 57221, "tokens": 3580887040, "source": "stack_v3", "source_tokens": {"stack_v3": 2685927424, "fineweb_edu": 894959616}, "loss": 0.8958183825016022, "ce_loss": 0.8948171436786652, "perplexity": 2.4468883195553763, "z_loss": 10.01253604888916, "grad_norm": 0.14482063055038452, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1184787.5403435018, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 22298.60802920023}
{"step": 6840, "total_steps": 57221, "tokens": 3586129920, "source": "stack_v3", "source_tokens": {"stack_v3": 2689597440, "fineweb_edu": 896532480}, "loss": 0.8986177295446396, "ce_loss": 0.8976914808154106, "perplexity": 2.4539316190812372, "z_loss": 9.262539744377136, "grad_norm": 0.16528719663619995, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1159757.7980635678, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 22775.33302565664}
{"step": 6850, "total_steps": 57221, "tokens": 3591372800, "source": "stack_v3", "source_tokens": {"stack_v3": 2693791744, "fineweb_edu": 897581056}, "loss": 0.9859590828418732, "ce_loss": 0.9848735108971596, "perplexity": 2.6774731918191272, "z_loss": 10.855822682380676, "grad_norm": 0.16226376593112946, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099615.706402611, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 24016.23316785437}
{"step": 6860, "total_steps": 57221, "tokens": 3596615680, "source": "stack_v3", "source_tokens": {"stack_v3": 2697461760, "fineweb_edu": 899153920}, "loss": 0.8856022879481316, "ce_loss": 0.8845306560397148, "perplexity": 2.421847445225842, "z_loss": 10.716272234916687, "grad_norm": 0.15617352724075317, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1123507.7305874324, "data_wait_seconds": 1.8656253814697266e-05, "eta_seconds": 23500.847925803628}
{"step": 6870, "total_steps": 57221, "tokens": 3601858560, "source": "stack_v3", "source_tokens": {"stack_v3": 2701656064, "fineweb_edu": 900202496}, "loss": 1.0248361825942993, "ce_loss": 1.0238735601305962, "perplexity": 2.783957732681909, "z_loss": 9.626208186149597, "grad_norm": 0.1321296989917755, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157145.3339973385, "data_wait_seconds": 1.6391277313232422e-05, "eta_seconds": 22813.15982047655}
{"step": 6880, "total_steps": 57221, "tokens": 3607101440, "source": "stack_v3", "source_tokens": {"stack_v3": 2705326080, "fineweb_edu": 901775360}, "loss": 0.9379316791892052, "ce_loss": 0.9367532134056091, "perplexity": 2.5516831835667797, "z_loss": 11.784496784210205, "grad_norm": 0.20320434868335724, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147325.8592993717, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 23003.83831330808}
{"step": 6890, "total_steps": 57221, "tokens": 3612344320, "source": "stack_v3", "source_tokens": {"stack_v3": 2709520384, "fineweb_edu": 902823936}, "loss": 0.9597009867429733, "ce_loss": 0.9587164744734764, "perplexity": 2.6083464447155413, "z_loss": 9.845086216926575, "grad_norm": 0.1123073622584343, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1086645.6347954583, "data_wait_seconds": 1.7464160919189453e-05, "eta_seconds": 24283.588720224332}
{"step": 6900, "total_steps": 57221, "tokens": 3617587200, "source": "stack_v3", "source_tokens": {"stack_v3": 2713190400, "fineweb_edu": 904396800}, "loss": 1.0271195620298386, "ce_loss": 1.026276782155037, "perplexity": 2.7906562470048524, "z_loss": 8.427820324897766, "grad_norm": 0.15434759855270386, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116667.0960391592, "data_wait_seconds": 2.5928020477294922e-05, "eta_seconds": 23626.03222892387}
{"step": 6910, "total_steps": 57221, "tokens": 3622830080, "source": "stack_v3", "source_tokens": {"stack_v3": 2717384704, "fineweb_edu": 905445376}, "loss": 1.1398093700408936, "ce_loss": 1.1389214396476746, "perplexity": 3.123397774815878, "z_loss": 8.879302740097046, "grad_norm": 0.13197582960128784, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099813.0315372928, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 23983.321858925978}
{"step": 6920, "total_steps": 57221, "tokens": 3628072960, "source": "stack_v3", "source_tokens": {"stack_v3": 2721054720, "fineweb_edu": 907018240}, "loss": 1.0604169070720673, "ce_loss": 1.059482902288437, "perplexity": 2.8848788392616114, "z_loss": 9.340114951133728, "grad_norm": 0.15394410490989685, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1160821.212699577, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 22718.33659782121}
{"step": 6930, "total_steps": 57221, "tokens": 3633315840, "source": "stack_v3", "source_tokens": {"stack_v3": 2725249024, "fineweb_edu": 908066816}, "loss": 0.9405986722558737, "ce_loss": 0.9395956490188837, "perplexity": 2.5589464965810436, "z_loss": 10.030103206634521, "grad_norm": 0.10979746282100677, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1129605.9106849213, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23341.489196009003}
{"step": 6940, "total_steps": 57221, "tokens": 3638558720, "source": "stack_v3", "source_tokens": {"stack_v3": 2728919040, "fineweb_edu": 909639680}, "loss": 0.9365738220512867, "ce_loss": 0.9354345872998238, "perplexity": 2.5483206849336932, "z_loss": 11.392480492591858, "grad_norm": 0.14790576696395874, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145309.460776925, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 23016.87201825576}
{"step": 6950, "total_steps": 57221, "tokens": 3643801600, "source": "stack_v3", "source_tokens": {"stack_v3": 2733113344, "fineweb_edu": 910688256}, "loss": 0.9861823432147503, "ce_loss": 0.9851538240909576, "perplexity": 2.678223828082544, "z_loss": 10.28504765033722, "grad_norm": 0.14205648005008698, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1096320.6221571183, "data_wait_seconds": 2.1159648895263672e-05, "eta_seconds": 24040.593479069652}
{"step": 6960, "total_steps": 57221, "tokens": 3649044480, "source": "stack_v3", "source_tokens": {"stack_v3": 2736783360, "fineweb_edu": 912261120}, "loss": 1.1883079260587692, "ce_loss": 1.1872120797634125, "perplexity": 3.277929849720767, "z_loss": 10.95863676071167, "grad_norm": 0.19410544633865356, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147033.094163901, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 22973.143193577882}
{"step": 6970, "total_steps": 57221, "tokens": 3654287360, "source": "stack_v3", "source_tokens": {"stack_v3": 2740977664, "fineweb_edu": 913309696}, "loss": 0.9024568982422352, "ce_loss": 0.9014209620654583, "perplexity": 2.4631005981836904, "z_loss": 10.359495401382446, "grad_norm": 0.09439988434314728, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1115565.3947015314, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 23616.466381201055}
{"step": 6980, "total_steps": 57221, "tokens": 3659530240, "source": "stack_v3", "source_tokens": {"stack_v3": 2744647680, "fineweb_edu": 914882560}, "loss": 1.022230502218008, "ce_loss": 1.0210596527904272, "perplexity": 2.776134945044755, "z_loss": 11.708315253257751, "grad_norm": 0.18485398590564728, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1097663.4369640118, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 23996.854475588774}
{"step": 6990, "total_steps": 57221, "tokens": 3664773120, "source": "stack_v3", "source_tokens": {"stack_v3": 2748841984, "fineweb_edu": 915931136}, "loss": 0.9174826890230179, "ce_loss": 0.9164294376969337, "perplexity": 2.50034678860719, "z_loss": 10.532458186149597, "grad_norm": 0.1050318256020546, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1132147.7684604677, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23261.298227714145}
{"step": 7000, "total_steps": 57221, "tokens": 3670016000, "source": "stack_v3", "source_tokens": {"stack_v3": 2752512000, "fineweb_edu": 917504000}, "loss": 1.055555872619152, "ce_loss": 1.054477535188198, "perplexity": 2.8704750398555885, "z_loss": 10.783336162567139, "grad_norm": 0.1607469916343689, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137988.6749599755, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 23137.298796867253}
{"step": 7010, "total_steps": 57221, "tokens": 3675258880, "source": "stack_v3", "source_tokens": {"stack_v3": 2756706304, "fineweb_edu": 918552576}, "loss": 0.9869434535503387, "ce_loss": 0.9857942461967468, "perplexity": 2.6799395711675573, "z_loss": 11.492019772529602, "grad_norm": 0.12796247005462646, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1102469.1666860103, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 23877.983997621803}
{"step": 7020, "total_steps": 57221, "tokens": 3680501760, "source": "stack_v3", "source_tokens": {"stack_v3": 2760376320, "fineweb_edu": 920125440}, "loss": 1.007334679365158, "ce_loss": 1.006185457110405, "perplexity": 2.735147752072558, "z_loss": 11.492130994796753, "grad_norm": 0.168076753616333, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1072603.2828493232, "data_wait_seconds": 2.3305416107177734e-05, "eta_seconds": 24537.961668440374}
{"step": 7030, "total_steps": 57221, "tokens": 3685744640, "source": "stack_v3", "source_tokens": {"stack_v3": 2764570624, "fineweb_edu": 921174016}, "loss": 0.8492768220603466, "ce_loss": 0.8480123579502106, "perplexity": 2.3350010900467866, "z_loss": 12.644594550132751, "grad_norm": 0.09823744744062424, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138742.0309236394, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 23108.179592401953}
{"step": 7040, "total_steps": 57221, "tokens": 3690987520, "source": "stack_v3", "source_tokens": {"stack_v3": 2768240640, "fineweb_edu": 922746880}, "loss": 0.8331454172730446, "ce_loss": 0.832070417702198, "perplexity": 2.2980717866786473, "z_loss": 10.750025272369385, "grad_norm": 0.1434181034564972, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1174084.8604783884, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 22408.10129284882}
{"step": 7050, "total_steps": 57221, "tokens": 3696230400, "source": "stack_v3", "source_tokens": {"stack_v3": 2772434944, "fineweb_edu": 923795456}, "loss": 0.7652379423379898, "ce_loss": 0.7641853615641594, "perplexity": 2.1472444341204078, "z_loss": 10.52569317817688, "grad_norm": 0.10544606298208237, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120673.244620507, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23471.39964861679}
{"step": 7060, "total_steps": 57221, "tokens": 3701473280, "source": "stack_v3", "source_tokens": {"stack_v3": 2776104960, "fineweb_edu": 925368320}, "loss": 0.9952572323381901, "ce_loss": 0.9943523779511452, "perplexity": 2.702973269237599, "z_loss": 9.048593759536743, "grad_norm": 0.25784364342689514, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135574.404822262, "data_wait_seconds": 2.282857894897461e-05, "eta_seconds": 23158.787841925863}
{"step": 7070, "total_steps": 57221, "tokens": 3706716160, "source": "stack_v3", "source_tokens": {"stack_v3": 2780299264, "fineweb_edu": 926416896}, "loss": 0.9226596653461456, "ce_loss": 0.9216810166835785, "perplexity": 2.513512096331018, "z_loss": 9.786613464355469, "grad_norm": 0.1290389597415924, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157940.7384687646, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 22706.93392717978}
{"step": 7080, "total_steps": 57221, "tokens": 3711959040, "source": "stack_v3", "source_tokens": {"stack_v3": 2783969280, "fineweb_edu": 927989760}, "loss": 0.7840400338172913, "ce_loss": 0.7831802293658257, "perplexity": 2.1884208909125196, "z_loss": 8.597966432571411, "grad_norm": 0.13995765149593353, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1115929.347729194, "data_wait_seconds": 2.09808349609375e-05, "eta_seconds": 23557.083621372236}
{"step": 7090, "total_steps": 57221, "tokens": 3717201920, "source": "stack_v3", "source_tokens": {"stack_v3": 2788163584, "fineweb_edu": 929038336}, "loss": 1.181264005601406, "ce_loss": 1.1804152205586433, "perplexity": 3.2557257665433186, "z_loss": 8.487961292266846, "grad_norm": 0.475241482257843, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122470.3558162772, "data_wait_seconds": 1.71661376953125e-05, "eta_seconds": 23415.137819730444}
{"step": 7100, "total_steps": 57221, "tokens": 3722444800, "source": "stack_v3", "source_tokens": {"stack_v3": 2791833600, "fineweb_edu": 930611200}, "loss": 0.9480971097946167, "ce_loss": 0.9471777081489563, "perplexity": 2.578422320351054, "z_loss": 9.193872928619385, "grad_norm": 0.14966130256652832, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128531.3023775527, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 23284.73755636136}
{"step": 7110, "total_steps": 57221, "tokens": 3727687680, "source": "stack_v3", "source_tokens": {"stack_v3": 2796027904, "fineweb_edu": 931659776}, "loss": 0.9490137733519077, "ce_loss": 0.947868425399065, "perplexity": 2.5802038963375877, "z_loss": 11.453360080718994, "grad_norm": 0.09680842608213425, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137180.4457462325, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 23103.028563562548}
{"step": 7120, "total_steps": 57221, "tokens": 3732930560, "source": "stack_v3", "source_tokens": {"stack_v3": 2799697920, "fineweb_edu": 933232640}, "loss": 0.9987564086914062, "ce_loss": 0.9976062253117561, "perplexity": 2.7117826561031464, "z_loss": 11.501946330070496, "grad_norm": 0.12340949475765228, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149944.9352949718, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 22842.023677648744}
{"step": 7130, "total_steps": 57221, "tokens": 3738173440, "source": "stack_v3", "source_tokens": {"stack_v3": 2803892224, "fineweb_edu": 934281216}, "loss": 0.9928839355707169, "ce_loss": 0.9917894676327705, "perplexity": 2.696054660829097, "z_loss": 10.944661974906921, "grad_norm": 0.10752008110284805, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1104566.4552451384, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 23775.687225782778}
{"step": 7140, "total_steps": 57221, "tokens": 3743416320, "source": "stack_v3", "source_tokens": {"stack_v3": 2807562240, "fineweb_edu": 935854080}, "loss": 0.9325434565544128, "ce_loss": 0.9316444993019104, "perplexity": 2.538680605205918, "z_loss": 8.989488124847412, "grad_norm": 0.16158434748649597, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1148818.9965488254, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 22855.283346530283}
{"step": 7150, "total_steps": 57221, "tokens": 3748659200, "source": "stack_v3", "source_tokens": {"stack_v3": 2811756544, "fineweb_edu": 936902656}, "loss": 0.9438107311725616, "ce_loss": 0.9429326727986336, "perplexity": 2.567500025618662, "z_loss": 8.780643463134766, "grad_norm": 0.12231194227933884, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154087.3813619642, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 22746.406575401776}
{"step": 7160, "total_steps": 57221, "tokens": 3753902080, "source": "stack_v3", "source_tokens": {"stack_v3": 2815426560, "fineweb_edu": 938475520}, "loss": 0.9598158150911331, "ce_loss": 0.9588623344898224, "perplexity": 2.6087269259184263, "z_loss": 9.534974694252014, "grad_norm": 0.1458936184644699, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128366.5558522197, "data_wait_seconds": 1.9669532775878906e-05, "eta_seconds": 23260.258631271776}
{"step": 7170, "total_steps": 57221, "tokens": 3759144960, "source": "stack_v3", "source_tokens": {"stack_v3": 2819620864, "fineweb_edu": 939524096}, "loss": 0.732900433242321, "ce_loss": 0.7318453192710876, "perplexity": 2.0789133291182953, "z_loss": 10.551137447357178, "grad_norm": 0.08528358489274979, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118219.9507510427, "data_wait_seconds": 1.913309097290039e-05, "eta_seconds": 23466.631070547042}
{"step": 7180, "total_steps": 57221, "tokens": 3764387840, "source": "stack_v3", "source_tokens": {"stack_v3": 2823290880, "fineweb_edu": 941096960}, "loss": 0.7016187235713005, "ce_loss": 0.7005519550293684, "perplexity": 2.01486451521098, "z_loss": 10.667619347572327, "grad_norm": 0.14662741124629974, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1152058.0917980985, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 22772.820525961695}
{"step": 7190, "total_steps": 57221, "tokens": 3769630720, "source": "stack_v3", "source_tokens": {"stack_v3": 2827485184, "fineweb_edu": 942145536}, "loss": 0.8485056720674038, "ce_loss": 0.847686942666769, "perplexity": 2.3342413686244177, "z_loss": 8.18712842464447, "grad_norm": 0.08179942518472672, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1136916.6399986104, "data_wait_seconds": 1.537799835205078e-05, "eta_seconds": 23071.49737911485}
{"step": 7200, "total_steps": 57221, "tokens": 3774873600, "source": "stack_v3", "source_tokens": {"stack_v3": 2831155200, "fineweb_edu": 943718400}, "loss": 1.143916692584753, "ce_loss": 1.1429950594902039, "perplexity": 3.1361472606158367, "z_loss": 9.216561913490295, "grad_norm": 0.18508601188659668, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141980.334751974, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 22964.604207213266}
{"step": 7210, "total_steps": 57221, "tokens": 3780116480, "source": "stack_v3", "source_tokens": {"stack_v3": 2835349504, "fineweb_edu": 944766976}, "loss": 0.9632534012198448, "ce_loss": 0.9626284688711166, "perplexity": 2.618570266121884, "z_loss": 6.249370992183685, "grad_norm": 0.1237792819738388, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098964.5632164308, "data_wait_seconds": 1.8537044525146484e-05, "eta_seconds": 23858.716102055274}
{"step": 7220, "total_steps": 57221, "tokens": 3785359360, "source": "stack_v3", "source_tokens": {"stack_v3": 2839019520, "fineweb_edu": 946339840}, "loss": 1.08203886449337, "ce_loss": 1.0809964388608932, "perplexity": 2.9476152065607932, "z_loss": 10.424510717391968, "grad_norm": 0.18103736639022827, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1095866.2833776583, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 23921.38624723605}
{"step": 7230, "total_steps": 57221, "tokens": 3790602240, "source": "stack_v3", "source_tokens": {"stack_v3": 2843213824, "fineweb_edu": 947388416}, "loss": 1.0527448505163193, "ce_loss": 1.0518015623092651, "perplexity": 2.8628039948310877, "z_loss": 9.432754397392273, "grad_norm": 0.13147501647472382, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145159.220260532, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 22887.121106213657}
{"step": 7240, "total_steps": 57221, "tokens": 3795845120, "source": "stack_v3", "source_tokens": {"stack_v3": 2846883840, "fineweb_edu": 948961280}, "loss": 0.9933454617857933, "ce_loss": 0.9921460598707199, "perplexity": 2.697016224427104, "z_loss": 11.994043469429016, "grad_norm": 0.14897724986076355, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157400.9457913132, "data_wait_seconds": 1.8358230590820312e-05, "eta_seconds": 22640.516214615895}
{"step": 7250, "total_steps": 57221, "tokens": 3801088000, "source": "stack_v3", "source_tokens": {"stack_v3": 2851078144, "fineweb_edu": 950009856}, "loss": 0.9628603160381317, "ce_loss": 0.961872786283493, "perplexity": 2.6165922056543653, "z_loss": 9.87526559829712, "grad_norm": 0.1491415947675705, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146758.7774627309, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 22846.053167316135}
{"step": 7260, "total_steps": 57221, "tokens": 3806330880, "source": "stack_v3", "source_tokens": {"stack_v3": 2854748160, "fineweb_edu": 951582720}, "loss": 1.1783738732337952, "ce_loss": 1.1774639636278152, "perplexity": 3.246131447923403, "z_loss": 9.098912954330444, "grad_norm": 0.20926053822040558, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1112554.5497849488, "data_wait_seconds": 2.2411346435546875e-05, "eta_seconds": 23543.71668792609}
{"step": 7270, "total_steps": 57221, "tokens": 3811573760, "source": "stack_v3", "source_tokens": {"stack_v3": 2858942464, "fineweb_edu": 952631296}, "loss": 1.2277697771787643, "ce_loss": 1.226675882935524, "perplexity": 3.4098758497528032, "z_loss": 10.93902051448822, "grad_norm": 0.13164886832237244, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1163138.256403844, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 22515.316726808203}
{"step": 7280, "total_steps": 57221, "tokens": 3816816640, "source": "stack_v3", "source_tokens": {"stack_v3": 2862612480, "fineweb_edu": 954204160}, "loss": 0.9604030475020409, "ce_loss": 0.9594870209693909, "perplexity": 2.610357071469865, "z_loss": 9.160181760787964, "grad_norm": 0.165270134806633, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143096.2889384173, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 22905.49240109604}
{"step": 7290, "total_steps": 57221, "tokens": 3822059520, "source": "stack_v3", "source_tokens": {"stack_v3": 2866806784, "fineweb_edu": 955252736}, "loss": 1.0419490486383438, "ce_loss": 1.0411133021116257, "perplexity": 2.8323685416033455, "z_loss": 8.357456922531128, "grad_norm": 0.1262313425540924, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1139249.8590728117, "data_wait_seconds": 1.823902130126953e-05, "eta_seconds": 22978.225778588327}
{"step": 7300, "total_steps": 57221, "tokens": 3827302400, "source": "stack_v3", "source_tokens": {"stack_v3": 2870476800, "fineweb_edu": 956825600}, "loss": 0.9145669341087341, "ce_loss": 0.9137310460209846, "perplexity": 2.4936089683731937, "z_loss": 8.359090566635132, "grad_norm": 0.15616805851459503, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1118427.4799008146, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 23401.336314019278}
{"step": 7310, "total_steps": 57221, "tokens": 3832545280, "source": "stack_v3", "source_tokens": {"stack_v3": 2874671104, "fineweb_edu": 957874176}, "loss": 0.9185099899768829, "ce_loss": 0.9176759421825409, "perplexity": 2.503465425388326, "z_loss": 8.340358257293701, "grad_norm": 0.3901281952857971, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108864.847319901, "data_wait_seconds": 1.4483928680419922e-05, "eta_seconds": 23598.416690046666}
{"step": 7320, "total_steps": 57221, "tokens": 3837788160, "source": "stack_v3", "source_tokens": {"stack_v3": 2878341120, "fineweb_edu": 959447040}, "loss": 0.7677846178412437, "ce_loss": 0.7669884338974953, "perplexity": 2.153271759154125, "z_loss": 7.961849093437195, "grad_norm": 0.16824646294116974, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150216.6273599067, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 22745.464826091196}
{"step": 7330, "total_steps": 57221, "tokens": 3843031040, "source": "stack_v3", "source_tokens": {"stack_v3": 2882535424, "fineweb_edu": 960495616}, "loss": 1.135110393166542, "ce_loss": 1.1341404169797897, "perplexity": 3.108500380022027, "z_loss": 9.699819207191467, "grad_norm": 0.1614328920841217, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1101687.074156984, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 23742.648501177555}
{"step": 7340, "total_steps": 57221, "tokens": 3848273920, "source": "stack_v3", "source_tokens": {"stack_v3": 2886205440, "fineweb_edu": 962068480}, "loss": 0.948038324713707, "ce_loss": 0.9469250291585922, "perplexity": 2.5777708895073252, "z_loss": 11.132924199104309, "grad_norm": 0.13892443478107452, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149514.412689292, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 22750.23765801941}
{"step": 7350, "total_steps": 57221, "tokens": 3853516800, "source": "stack_v3", "source_tokens": {"stack_v3": 2890399744, "fineweb_edu": 963117056}, "loss": 1.0728279054164886, "ce_loss": 1.0718837529420853, "perplexity": 2.9208765305856352, "z_loss": 9.441495418548584, "grad_norm": 0.13352848589420319, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1100141.8440313532, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 23766.465517018227}
{"step": 7360, "total_steps": 57221, "tokens": 3858759680, "source": "stack_v3", "source_tokens": {"stack_v3": 2894069760, "fineweb_edu": 964689920}, "loss": 1.505151391029358, "ce_loss": 1.504196047782898, "perplexity": 4.500533961206698, "z_loss": 9.553225874900818, "grad_norm": 0.5768117904663086, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1147496.4285540222, "data_wait_seconds": 2.1696090698242188e-05, "eta_seconds": 22781.108219169775}
{"step": 7370, "total_steps": 57221, "tokens": 3864002560, "source": "stack_v3", "source_tokens": {"stack_v3": 2898264064, "fineweb_edu": 965738496}, "loss": 1.1212237924337387, "ce_loss": 1.120275691151619, "perplexity": 3.0656992729616146, "z_loss": 9.480971217155457, "grad_norm": 0.2510536015033722, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1127153.4313063119, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 23187.612896417966}
{"step": 7380, "total_steps": 57221, "tokens": 3869245440, "source": "stack_v3", "source_tokens": {"stack_v3": 2901934080, "fineweb_edu": 967311360}, "loss": 1.1013706624507904, "ce_loss": 1.100190982222557, "perplexity": 3.004739821041395, "z_loss": 11.796749830245972, "grad_norm": 0.15289607644081116, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116659.4409825322, "data_wait_seconds": 2.0623207092285156e-05, "eta_seconds": 23400.82714655413}
{"step": 7390, "total_steps": 57221, "tokens": 3874488320, "source": "stack_v3", "source_tokens": {"stack_v3": 2906128384, "fineweb_edu": 968359936}, "loss": 1.1789182424545288, "ce_loss": 1.1777511239051819, "perplexity": 3.2470637417828323, "z_loss": 11.67122483253479, "grad_norm": 0.12273520231246948, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150383.8446342223, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 22710.256060929736}
{"step": 7400, "total_steps": 57221, "tokens": 3879731200, "source": "stack_v3", "source_tokens": {"stack_v3": 2909798400, "fineweb_edu": 969932800}, "loss": 1.0897046476602554, "ce_loss": 1.088637575507164, "perplexity": 2.9702246078895103, "z_loss": 10.670858979225159, "grad_norm": 0.16964556276798248, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1168131.59065171, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 22360.724604175193}
{"step": 7410, "total_steps": 57221, "tokens": 3884974080, "source": "stack_v3", "source_tokens": {"stack_v3": 2913992704, "fineweb_edu": 970981376}, "loss": 1.1577285826206207, "ce_loss": 1.1567639857530594, "perplexity": 3.1796272907391434, "z_loss": 9.645741939544678, "grad_norm": 0.11123043298721313, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166860.2952693098, "data_wait_seconds": 2.0265579223632812e-05, "eta_seconds": 22380.593483106466}
{"step": 7420, "total_steps": 57221, "tokens": 3890216960, "source": "stack_v3", "source_tokens": {"stack_v3": 2917662720, "fineweb_edu": 972554240}, "loss": 1.1924550235271454, "ce_loss": 1.191442757844925, "perplexity": 3.2918270923401187, "z_loss": 10.122538208961487, "grad_norm": 0.19014595448970795, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122883.4932367539, "data_wait_seconds": 1.7583370208740234e-05, "eta_seconds": 23252.44176912564}
{"step": 7430, "total_steps": 57221, "tokens": 3895459840, "source": "stack_v3", "source_tokens": {"stack_v3": 2921857024, "fineweb_edu": 973602816}, "loss": 1.3482950031757355, "ce_loss": 1.3471913933753967, "perplexity": 3.8466067397732697, "z_loss": 11.036067187786102, "grad_norm": 0.21323399245738983, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141920.1438876865, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 22860.215138272848}
{"step": 7440, "total_steps": 57221, "tokens": 3900702720, "source": "stack_v3", "source_tokens": {"stack_v3": 2925527040, "fineweb_edu": 975175680}, "loss": 1.1002169102430344, "ce_loss": 1.0993672013282776, "perplexity": 3.002265593035341, "z_loss": 8.49709540605545, "grad_norm": 0.16984222829341888, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1154036.56654084, "data_wait_seconds": 1.913309097290039e-05, "eta_seconds": 22615.658841930097}
{"step": 7450, "total_steps": 57221, "tokens": 3905945600, "source": "stack_v3", "source_tokens": {"stack_v3": 2929721344, "fineweb_edu": 976224256}, "loss": 1.3064040541648865, "ce_loss": 1.305361032485962, "perplexity": 3.689020709629811, "z_loss": 10.430259227752686, "grad_norm": 0.17316217720508575, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1117460.1043531038, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 23351.217907780087}
{"step": 7460, "total_steps": 57221, "tokens": 3911188480, "source": "stack_v3", "source_tokens": {"stack_v3": 2933391360, "fineweb_edu": 977797120}, "loss": 0.9705339521169662, "ce_loss": 0.9697552621364594, "perplexity": 2.6372989334585677, "z_loss": 7.787117600440979, "grad_norm": 0.19153477251529694, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1110026.8566880631, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 23502.86514493893}
{"step": 7470, "total_steps": 57221, "tokens": 3916431360, "source": "stack_v3", "source_tokens": {"stack_v3": 2937585664, "fineweb_edu": 978845696}, "loss": 1.08075363188982, "ce_loss": 1.0798331648111343, "perplexity": 2.9441883158752704, "z_loss": 9.204587459564209, "grad_norm": 0.13799135386943817, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 927618.3768377111, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 28118.857162920758}
{"step": 7480, "total_steps": 57221, "tokens": 3921674240, "source": "stack_v3", "source_tokens": {"stack_v3": 2941255680, "fineweb_edu": 980418560}, "loss": 0.918383814394474, "ce_loss": 0.9174229763448238, "perplexity": 2.5028322142535666, "z_loss": 9.608402609825134, "grad_norm": 0.11344647407531738, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 996709.9200741374, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 26164.408756020246}
{"step": 7490, "total_steps": 57221, "tokens": 3926917120, "source": "stack_v3", "source_tokens": {"stack_v3": 2945449984, "fineweb_edu": 981467136}, "loss": 1.0204894039779902, "ce_loss": 1.0196629371494055, "perplexity": 2.772260180547426, "z_loss": 8.2645902633667, "grad_norm": 0.095770463347435, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1064486.4057723114, "data_wait_seconds": 1.7344951629638672e-05, "eta_seconds": 24493.57994485926}
{"step": 7500, "total_steps": 57221, "tokens": 3932160000, "source": "stack_v3", "source_tokens": {"stack_v3": 2949120000, "fineweb_edu": 983040000}, "loss": 0.9792988933622837, "ce_loss": 0.9786383397877216, "perplexity": 2.6608306268655073, "z_loss": 6.605465471744537, "grad_norm": 0.15125969052314758, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130962.333104846, "data_wait_seconds": 1.6510486602783203e-05, "eta_seconds": 23049.255697522312}
{"step": 7510, "total_steps": 57221, "tokens": 3937402880, "source": "stack_v3", "source_tokens": {"stack_v3": 2953314304, "fineweb_edu": 984088576}, "loss": 1.0949489921331406, "ce_loss": 1.0940929502248764, "perplexity": 2.9864725751999113, "z_loss": 8.560572147369385, "grad_norm": 0.1285250037908554, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130446.0580657579, "data_wait_seconds": 1.728534698486328e-05, "eta_seconds": 23055.144413165748}
{"step": 7520, "total_steps": 57221, "tokens": 3942645760, "source": "stack_v3", "source_tokens": {"stack_v3": 2956984320, "fineweb_edu": 985661440}, "loss": 1.0514015704393387, "ce_loss": 1.0504122972488403, "perplexity": 2.858829562675329, "z_loss": 9.892766952514648, "grad_norm": 0.14640763401985168, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1153443.1095784018, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 22590.931467373623}
{"step": 7530, "total_steps": 57221, "tokens": 3947888640, "source": "stack_v3", "source_tokens": {"stack_v3": 2961178624, "fineweb_edu": 986710016}, "loss": 1.2615333050489426, "ce_loss": 1.2605434209108353, "perplexity": 3.5273377957549914, "z_loss": 9.898979187011719, "grad_norm": 0.12714260816574097, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1150070.269414317, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 22652.625715876697}
{"step": 7540, "total_steps": 57221, "tokens": 3953131520, "source": "stack_v3", "source_tokens": {"stack_v3": 2964848640, "fineweb_edu": 988282880}, "loss": 1.1853631436824799, "ce_loss": 1.1845354735851288, "perplexity": 3.26916785387021, "z_loss": 8.27651333808899, "grad_norm": 0.1832597404718399, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1047630.7528693298, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 24862.642117617186}
{"step": 7550, "total_steps": 57221, "tokens": 3958374400, "source": "stack_v3", "source_tokens": {"stack_v3": 2969042944, "fineweb_edu": 989331456}, "loss": 1.0173417180776596, "ce_loss": 1.0163592547178268, "perplexity": 2.763116625310889, "z_loss": 9.824719786643982, "grad_norm": 0.11011070758104324, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131428.9557607723, "data_wait_seconds": 1.9550323486328125e-05, "eta_seconds": 23016.580464382427}
{"step": 7560, "total_steps": 57221, "tokens": 3963617280, "source": "stack_v3", "source_tokens": {"stack_v3": 2972712960, "fineweb_edu": 990904320}, "loss": 1.0088300853967667, "ce_loss": 1.0079280585050583, "perplexity": 2.7399181796295795, "z_loss": 9.020419716835022, "grad_norm": 0.14553077518939972, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144477.103128527, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 22749.58812965965}
{"step": 7570, "total_steps": 57221, "tokens": 3968860160, "source": "stack_v3", "source_tokens": {"stack_v3": 2976907264, "fineweb_edu": 991952896}, "loss": 1.22410848736763, "ce_loss": 1.2229792326688766, "perplexity": 3.397294000958741, "z_loss": 11.292346358299255, "grad_norm": 0.11201605945825577, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1084872.895362427, "data_wait_seconds": 2.1576881408691406e-05, "eta_seconds": 23994.644857731186}
{"step": 7580, "total_steps": 57221, "tokens": 3974103040, "source": "stack_v3", "source_tokens": {"stack_v3": 2980577280, "fineweb_edu": 993525760}, "loss": 1.112325981259346, "ce_loss": 1.1111679375171661, "perplexity": 3.037904405801834, "z_loss": 11.580374717712402, "grad_norm": 0.12560544908046722, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1160269.977327752, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 22430.897522610147}
{"step": 7590, "total_steps": 57221, "tokens": 3979345920, "source": "stack_v3", "source_tokens": {"stack_v3": 2984771584, "fineweb_edu": 994574336}, "loss": 0.9893143177032471, "ce_loss": 0.9880459308624268, "perplexity": 2.685980748866058, "z_loss": 12.68408989906311, "grad_norm": 0.09619423747062683, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1089587.8792890115, "data_wait_seconds": 1.6689300537109375e-05, "eta_seconds": 23881.18900237698}
{"step": 7600, "total_steps": 57221, "tokens": 3984588800, "source": "stack_v3", "source_tokens": {"stack_v3": 2988441600, "fineweb_edu": 996147200}, "loss": 1.030642792582512, "ce_loss": 1.0295935720205307, "perplexity": 2.799927634485396, "z_loss": 10.492228984832764, "grad_norm": 0.15037788450717926, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146498.937688989, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 22691.17776283296}
{"step": 7610, "total_steps": 57221, "tokens": 3989831680, "source": "stack_v3", "source_tokens": {"stack_v3": 2992635904, "fineweb_edu": 997195776}, "loss": 0.9483576193451881, "ce_loss": 0.9474094733595848, "perplexity": 2.5790199781986924, "z_loss": 9.481430172920227, "grad_norm": 0.19838331639766693, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1123329.0685177, "data_wait_seconds": 1.6927719116210938e-05, "eta_seconds": 23154.540418260494}
{"step": 7620, "total_steps": 57221, "tokens": 3995074560, "source": "stack_v3", "source_tokens": {"stack_v3": 2996305920, "fineweb_edu": 998768640}, "loss": 1.061035767197609, "ce_loss": 1.0600689351558685, "perplexity": 2.8865699685602535, "z_loss": 9.668192148208618, "grad_norm": 0.16941092908382416, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 397479.01834602037, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 65424.64945246931}
{"step": 7630, "total_steps": 57221, "tokens": 4000317440, "source": "stack_v3", "source_tokens": {"stack_v3": 3000500224, "fineweb_edu": 999817216}, "loss": 1.03147491812706, "ce_loss": 1.0304934531450272, "perplexity": 2.8024483705249117, "z_loss": 9.814769625663757, "grad_norm": 0.10909111052751541, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1096122.9089966912, "data_wait_seconds": 2.0325183868408203e-05, "eta_seconds": 23719.678100513527}
{"step": 7640, "total_steps": 57221, "tokens": 4005560320, "source": "stack_v3", "source_tokens": {"stack_v3": 3004170240, "fineweb_edu": 1001390080}, "loss": 1.066315844655037, "ce_loss": 1.0652846544981003, "perplexity": 2.9016648384607198, "z_loss": 10.312039613723755, "grad_norm": 0.14855213463306427, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1092082.3010899154, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 23802.638000869658}
{"step": 7650, "total_steps": 57221, "tokens": 4010803200, "source": "stack_v3", "source_tokens": {"stack_v3": 3008364544, "fineweb_edu": 1002438656}, "loss": 0.9448273032903671, "ce_loss": 0.9439704418182373, "perplexity": 2.570165880634702, "z_loss": 8.568529844284058, "grad_norm": 0.10365036129951477, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1093059.4617323135, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 23776.562675567107}
{"step": 7660, "total_steps": 57221, "tokens": 4016046080, "source": "stack_v3", "source_tokens": {"stack_v3": 3012034560, "fineweb_edu": 1004011520}, "loss": 0.8395715653896332, "ce_loss": 0.8386319875717163, "perplexity": 2.3132003244721013, "z_loss": 9.39598298072815, "grad_norm": 0.15411332249641418, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1159589.9848933232, "data_wait_seconds": 2.002716064453125e-05, "eta_seconds": 22407.880594441667}
{"step": 7670, "total_steps": 57221, "tokens": 4021288960, "source": "stack_v3", "source_tokens": {"stack_v3": 3016228864, "fineweb_edu": 1005060096}, "loss": 1.1384711265563965, "ce_loss": 1.1373955011367798, "perplexity": 3.118635296415157, "z_loss": 10.756321668624878, "grad_norm": 0.1218758150935173, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098347.708709081, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 23652.53811157262}
{"step": 7680, "total_steps": 57221, "tokens": 4026531840, "source": "stack_v3", "source_tokens": {"stack_v3": 3019898880, "fineweb_edu": 1006632960}, "loss": 1.0489757806062698, "ce_loss": 1.04799123108387, "perplexity": 2.8519165189945634, "z_loss": 9.845552921295166, "grad_norm": 0.16003596782684326, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1115775.3930603259, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 23278.402016700245}
{"step": 7690, "total_steps": 57221, "tokens": 4031774720, "source": "stack_v3", "source_tokens": {"stack_v3": 3024093184, "fineweb_edu": 1007681536}, "loss": 1.1506891250610352, "ce_loss": 1.149765208363533, "perplexity": 3.1574514794524386, "z_loss": 9.23909044265747, "grad_norm": 0.12254491448402405, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1159985.1938882351, "data_wait_seconds": 1.8417835235595703e-05, "eta_seconds": 22386.68684464437}
{"step": 7700, "total_steps": 57221, "tokens": 4037017600, "source": "stack_v3", "source_tokens": {"stack_v3": 3027763200, "fineweb_edu": 1009254400}, "loss": 1.163941502571106, "ce_loss": 1.1629746556282043, "perplexity": 3.199436356277631, "z_loss": 9.668557405471802, "grad_norm": 0.16568824648857117, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 986215.8421777138, "data_wait_seconds": 2.092123031616211e-05, "eta_seconds": 26325.862239922862}
{"step": 7710, "total_steps": 57221, "tokens": 4042260480, "source": "stack_v3", "source_tokens": {"stack_v3": 3031957504, "fineweb_edu": 1010302976}, "loss": 0.9679257273674011, "ce_loss": 0.9672021567821503, "perplexity": 2.6305742195350375, "z_loss": 7.23559296131134, "grad_norm": 0.1130964457988739, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1127711.983021832, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 23018.057722897734}
{"step": 7720, "total_steps": 57221, "tokens": 4047503360, "source": "stack_v3", "source_tokens": {"stack_v3": 3035627520, "fineweb_edu": 1011875840}, "loss": 1.2471331506967545, "ce_loss": 1.2463357746601105, "perplexity": 3.4775769573862934, "z_loss": 7.973613739013672, "grad_norm": 0.1711457073688507, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1046383.2450413783, "data_wait_seconds": 1.6570091247558594e-05, "eta_seconds": 24802.094990515383}
{"step": 7730, "total_steps": 57221, "tokens": 4052746240, "source": "stack_v3", "source_tokens": {"stack_v3": 3039821824, "fineweb_edu": 1012924416}, "loss": 1.0953525006771088, "ce_loss": 1.0944793820381165, "perplexity": 2.9876268662253604, "z_loss": 8.73129916191101, "grad_norm": 0.10037688165903091, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1173450.8227927098, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 22111.922592757503}
{"step": 7740, "total_steps": 57221, "tokens": 4057989120, "source": "stack_v3", "source_tokens": {"stack_v3": 3043491840, "fineweb_edu": 1014497280}, "loss": 1.1175909489393234, "ce_loss": 1.1166194528341293, "perplexity": 3.0545108121135613, "z_loss": 9.715099811553955, "grad_norm": 0.12628720700740814, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125203.5120878532, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 23055.394514245447}
{"step": 7750, "total_steps": 57221, "tokens": 4063232000, "source": "stack_v3", "source_tokens": {"stack_v3": 3047686144, "fineweb_edu": 1015545856}, "loss": 1.0055961906909943, "ce_loss": 1.0046533942222595, "perplexity": 2.730960542061569, "z_loss": 9.427813649177551, "grad_norm": 0.1253388375043869, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1046853.6337728644, "data_wait_seconds": 1.901388168334961e-05, "eta_seconds": 24775.925844116136}
{"step": 7760, "total_steps": 57221, "tokens": 4068474880, "source": "stack_v3", "source_tokens": {"stack_v3": 3051356160, "fineweb_edu": 1017118720}, "loss": 0.8423291221261024, "ce_loss": 0.8415163829922676, "perplexity": 2.319882140762402, "z_loss": 8.127357006072998, "grad_norm": 0.1342763453722, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1145730.9887267307, "data_wait_seconds": 2.378225326538086e-05, "eta_seconds": 22633.1707662181}
{"step": 7770, "total_steps": 57221, "tokens": 4073717760, "source": "stack_v3", "source_tokens": {"stack_v3": 3055550464, "fineweb_edu": 1018167296}, "loss": 0.9385003447532654, "ce_loss": 0.9377937763929367, "perplexity": 2.554339752566968, "z_loss": 7.065829575061798, "grad_norm": 0.10601746290922165, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125821.1678108042, "data_wait_seconds": 2.3424625396728516e-05, "eta_seconds": 23028.77489007823}
{"step": 7780, "total_steps": 57221, "tokens": 4078960640, "source": "stack_v3", "source_tokens": {"stack_v3": 3059220480, "fineweb_edu": 1019740160}, "loss": 1.1554565727710724, "ce_loss": 1.154460072517395, "perplexity": 3.1723101376191205, "z_loss": 9.964908123016357, "grad_norm": 0.1555827409029007, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1123073.6569450237, "data_wait_seconds": 2.193450927734375e-05, "eta_seconds": 23080.444634869462}
{"step": 7790, "total_steps": 57221, "tokens": 4084203520, "source": "stack_v3", "source_tokens": {"stack_v3": 3063414784, "fineweb_edu": 1020788736}, "loss": 1.0223764926195145, "ce_loss": 1.0213744342327118, "perplexity": 2.7770089583611024, "z_loss": 10.020326018333435, "grad_norm": 0.09096482396125793, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1095379.145642193, "data_wait_seconds": 2.1219253540039062e-05, "eta_seconds": 23659.20200608368}
{"step": 7800, "total_steps": 57221, "tokens": 4089446400, "source": "stack_v3", "source_tokens": {"stack_v3": 3067084800, "fineweb_edu": 1022361600}, "loss": 1.0379265546798706, "ce_loss": 1.036922812461853, "perplexity": 2.8205243643287297, "z_loss": 10.0374436378479, "grad_norm": 0.25425294041633606, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1094805.333349265, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23666.813460557023}
{"step": 7810, "total_steps": 57221, "tokens": 4094689280, "source": "stack_v3", "source_tokens": {"stack_v3": 3071279104, "fineweb_edu": 1023410176}, "loss": 0.9162155985832214, "ce_loss": 0.9155501276254654, "perplexity": 2.4981491748273665, "z_loss": 6.654848337173462, "grad_norm": 0.16331195831298828, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1139974.5691169796, "data_wait_seconds": 2.2470951080322266e-05, "eta_seconds": 22724.463704542257}
{"step": 7820, "total_steps": 57221, "tokens": 4099932160, "source": "stack_v3", "source_tokens": {"stack_v3": 3074949120, "fineweb_edu": 1024983040}, "loss": 0.9908002465963364, "ce_loss": 0.9898529797792435, "perplexity": 2.6908388355470474, "z_loss": 9.472561836242676, "grad_norm": 0.1817430704832077, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1117667.8627900588, "data_wait_seconds": 2.0623207092285156e-05, "eta_seconds": 23173.31355966977}
{"step": 7830, "total_steps": 57221, "tokens": 4105175040, "source": "stack_v3", "source_tokens": {"stack_v3": 3079143424, "fineweb_edu": 1026031616}, "loss": 1.0489931553602219, "ce_loss": 1.048124060034752, "perplexity": 2.852295361233836, "z_loss": 8.690683603286743, "grad_norm": 0.09929947555065155, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1115977.7681472357, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 23203.710413506717}
{"step": 7840, "total_steps": 57221, "tokens": 4110417920, "source": "stack_v3", "source_tokens": {"stack_v3": 3082813440, "fineweb_edu": 1027604480}, "loss": 1.0973969250917435, "ce_loss": 1.096530795097351, "perplexity": 2.9937620137041416, "z_loss": 8.661223769187927, "grad_norm": 0.14676471054553986, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116746.8282755073, "data_wait_seconds": 2.2232532501220703e-05, "eta_seconds": 23183.036140768785}
{"step": 7850, "total_steps": 57221, "tokens": 4115660800, "source": "stack_v3", "source_tokens": {"stack_v3": 3087007744, "fineweb_edu": 1028653056}, "loss": 1.054013192653656, "ce_loss": 1.0531385093927383, "perplexity": 2.8666339719504186, "z_loss": 8.746992766857147, "grad_norm": 0.11050337553024292, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1116899.1791268936, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 23175.179715177514}
{"step": 7860, "total_steps": 57221, "tokens": 4120903680, "source": "stack_v3", "source_tokens": {"stack_v3": 3090677760, "fineweb_edu": 1030225920}, "loss": 1.0757564455270767, "ce_loss": 1.0748243927955627, "perplexity": 2.9294784178461732, "z_loss": 9.320559024810791, "grad_norm": 0.1663648784160614, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144266.940111138, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 22616.310419216054}
{"step": 7870, "total_steps": 57221, "tokens": 4126146560, "source": "stack_v3", "source_tokens": {"stack_v3": 3094872064, "fineweb_edu": 1031274496}, "loss": 0.9774636626243591, "ce_loss": 0.9766609817743301, "perplexity": 2.6555744105265724, "z_loss": 8.026937246322632, "grad_norm": 0.1432133913040161, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1165117.6552676377, "data_wait_seconds": 1.627206802368164e-05, "eta_seconds": 22207.073528601322}
{"step": 7880, "total_steps": 57221, "tokens": 4131389440, "source": "stack_v3", "source_tokens": {"stack_v3": 3098542080, "fineweb_edu": 1032847360}, "loss": 1.0895598381757736, "ce_loss": 1.0887228697538376, "perplexity": 2.970477961764552, "z_loss": 8.369571328163147, "grad_norm": 0.18435877561569214, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1082000.4779409792, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 23908.13228588156}
{"step": 7890, "total_steps": 57221, "tokens": 4136632320, "source": "stack_v3", "source_tokens": {"stack_v3": 3102736384, "fineweb_edu": 1033895936}, "loss": 1.052179902791977, "ce_loss": 1.051309660077095, "perplexity": 2.8613961214521817, "z_loss": 8.702368974685669, "grad_norm": 0.10699079185724258, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1130589.8467791898, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 22875.99499825621}
{"step": 7900, "total_steps": 57221, "tokens": 4141875200, "source": "stack_v3", "source_tokens": {"stack_v3": 3106406400, "fineweb_edu": 1035468800}, "loss": 1.0340906381607056, "ce_loss": 1.0330411046743393, "perplexity": 2.8095971348154323, "z_loss": 10.495333194732666, "grad_norm": 0.3060981035232544, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1149040.501526055, "data_wait_seconds": 1.5735626220703125e-05, "eta_seconds": 22504.102131872205}
{"step": 7910, "total_steps": 57221, "tokens": 4147118080, "source": "stack_v3", "source_tokens": {"stack_v3": 3110600704, "fineweb_edu": 1036517376}, "loss": 0.9177932739257812, "ce_loss": 0.9168875068426132, "perplexity": 2.5014923826851687, "z_loss": 9.057735085487366, "grad_norm": 0.09438510239124298, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1097118.3737554897, "data_wait_seconds": 1.9490718841552734e-05, "eta_seconds": 23564.35051899124}
{"step": 7920, "total_steps": 57221, "tokens": 4152360960, "source": "stack_v3", "source_tokens": {"stack_v3": 3114270720, "fineweb_edu": 1038090240}, "loss": 1.0961519330739975, "ce_loss": 1.0952544212341309, "perplexity": 2.9899432916940634, "z_loss": 8.975232362747192, "grad_norm": 0.17449481785297394, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125679.857143062, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 22961.802928232588}
{"step": 7930, "total_steps": 57221, "tokens": 4157603840, "source": "stack_v3", "source_tokens": {"stack_v3": 3118465024, "fineweb_edu": 1039138816}, "loss": 1.191006451845169, "ce_loss": 1.1899836659431458, "perplexity": 3.287027516450289, "z_loss": 10.227998852729797, "grad_norm": 0.12209225445985794, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1098063.3371374826, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 23534.52235949156}
{"step": 7940, "total_steps": 57221, "tokens": 4162846720, "source": "stack_v3", "source_tokens": {"stack_v3": 3122135040, "fineweb_edu": 1040711680}, "loss": 1.0289100408554077, "ce_loss": 1.0281974971294403, "perplexity": 2.7960214531133247, "z_loss": 7.125492811203003, "grad_norm": 0.16907961666584015, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1187522.2890565011, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 21757.194385402137}
{"step": 7950, "total_steps": 57221, "tokens": 4168089600, "source": "stack_v3", "source_tokens": {"stack_v3": 3126329344, "fineweb_edu": 1041760256}, "loss": 1.1747980564832687, "ce_loss": 1.173917293548584, "perplexity": 3.234638882859078, "z_loss": 8.807803750038147, "grad_norm": 0.10529769957065582, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1113733.6726557116, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 23193.974496976007}
{"step": 7960, "total_steps": 57221, "tokens": 4173332480, "source": "stack_v3", "source_tokens": {"stack_v3": 3129999360, "fineweb_edu": 1043333120}, "loss": 1.1269142031669617, "ce_loss": 1.1259913742542267, "perplexity": 3.0832720107536535, "z_loss": 9.22841501235962, "grad_norm": 0.14949241280555725, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1107960.8820645805, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 23310.089677421143}
{"step": 7970, "total_steps": 57221, "tokens": 4178575360, "source": "stack_v3", "source_tokens": {"stack_v3": 3134193664, "fineweb_edu": 1044381696}, "loss": 1.504499763250351, "ce_loss": 1.503209263086319, "perplexity": 4.496095093630836, "z_loss": 12.905023455619812, "grad_norm": 0.13978011906147003, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131037.2554033075, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 22829.862161165103}
{"step": 7980, "total_steps": 57221, "tokens": 4183818240, "source": "stack_v3", "source_tokens": {"stack_v3": 3137863680, "fineweb_edu": 1045954560}, "loss": 1.0732320547103882, "ce_loss": 1.0721223801374435, "perplexity": 2.921573614328378, "z_loss": 11.096595525741577, "grad_norm": 0.18181195855140686, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120124.1524289984, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 23047.607449600473}
{"step": 7990, "total_steps": 57221, "tokens": 4189061120, "source": "stack_v3", "source_tokens": {"stack_v3": 3142057984, "fineweb_edu": 1047003136}, "loss": 0.8581227958202362, "ce_loss": 0.8572799265384674, "perplexity": 2.3567414573725918, "z_loss": 8.428706407546997, "grad_norm": 0.12312573194503784, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1094899.1461430432, "data_wait_seconds": 2.0384788513183594e-05, "eta_seconds": 23573.804921597708}
{"step": 8000, "total_steps": 57221, "tokens": 4194304000, "source": "stack_v3", "source_tokens": {"stack_v3": 3145728000, "fineweb_edu": 1048576000}, "loss": 1.2900590002536774, "ce_loss": 1.2890228629112244, "perplexity": 3.6292385589954903, "z_loss": 10.361164450645447, "grad_norm": 0.2229231297969818, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1110696.7326173128, "data_wait_seconds": 1.6748905181884766e-05, "eta_seconds": 23233.791225071764}
{"step": 8010, "total_steps": 57221, "tokens": 4199546880, "source": "stack_v3", "source_tokens": {"stack_v3": 3149922304, "fineweb_edu": 1049624576}, "loss": 1.132418155670166, "ce_loss": 1.1315925568342209, "perplexity": 3.1005904367848567, "z_loss": 8.256052494049072, "grad_norm": 0.47188201546669006, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1083022.448543557, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 23822.63927649544}
{"step": 8020, "total_steps": 57221, "tokens": 4204789760, "source": "stack_v3", "source_tokens": {"stack_v3": 3153592320, "fineweb_edu": 1051197440}, "loss": 1.0198642909526825, "ce_loss": 1.0189015567302704, "perplexity": 2.770150239265126, "z_loss": 9.627337098121643, "grad_norm": 0.13262924551963806, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1031916.0688758916, "data_wait_seconds": 1.728534698486328e-05, "eta_seconds": 24997.391762781423}
{"step": 8030, "total_steps": 57221, "tokens": 4210032640, "source": "stack_v3", "source_tokens": {"stack_v3": 3157786624, "fineweb_edu": 1052246016}, "loss": 1.3021908104419708, "ce_loss": 1.300930768251419, "perplexity": 3.672713522365784, "z_loss": 12.60069227218628, "grad_norm": 0.23974700272083282, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1127238.019532947, "data_wait_seconds": 2.1338462829589844e-05, "eta_seconds": 22878.901272941148}
{"step": 8040, "total_steps": 57221, "tokens": 4215275520, "source": "stack_v3", "source_tokens": {"stack_v3": 3161456640, "fineweb_edu": 1053818880}, "loss": 1.0003492087125778, "ce_loss": 0.9992818087339401, "perplexity": 2.7163302830664686, "z_loss": 10.674134731292725, "grad_norm": 0.1621863842010498, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1100924.609466019, "data_wait_seconds": 2.6285648345947266e-05, "eta_seconds": 23420.972025056602}
{"step": 8050, "total_steps": 57221, "tokens": 4220518400, "source": "stack_v3", "source_tokens": {"stack_v3": 3165650944, "fineweb_edu": 1054867456}, "loss": 1.0410830080509186, "ce_loss": 1.0398968011140823, "perplexity": 2.828925057372736, "z_loss": 11.861855626106262, "grad_norm": 0.10388703644275665, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1122280.3390669564, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 22970.62570073409}
{"step": 8060, "total_steps": 57221, "tokens": 4225761280, "source": "stack_v3", "source_tokens": {"stack_v3": 3169320960, "fineweb_edu": 1056440320}, "loss": 1.169074073433876, "ce_loss": 1.1680755913257599, "perplexity": 3.215798170202557, "z_loss": 9.984508514404297, "grad_norm": 0.39938366413116455, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131734.9511732813, "data_wait_seconds": 1.8715858459472656e-05, "eta_seconds": 22774.094493838493}
{"step": 8070, "total_steps": 57221, "tokens": 4231004160, "source": "stack_v3", "source_tokens": {"stack_v3": 3173515264, "fineweb_edu": 1057488896}, "loss": 1.0231836587190628, "ce_loss": 1.0219622254371643, "perplexity": 2.7786417396217344, "z_loss": 12.214311838150024, "grad_norm": 0.1098812073469162, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125230.2274230495, "data_wait_seconds": 1.9371509552001953e-05, "eta_seconds": 22901.08745035669}
{"step": 8080, "total_steps": 57221, "tokens": 4236247040, "source": "stack_v3", "source_tokens": {"stack_v3": 3177185280, "fineweb_edu": 1059061760}, "loss": 0.9346175789833069, "ce_loss": 0.9330503046512604, "perplexity": 2.542252005739819, "z_loss": 15.672682642936707, "grad_norm": 0.1710399091243744, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1141549.5910071968, "data_wait_seconds": 2.0802021026611328e-05, "eta_seconds": 22569.105330998776}
{"step": 8090, "total_steps": 57221, "tokens": 4241489920, "source": "stack_v3", "source_tokens": {"stack_v3": 3181379584, "fineweb_edu": 1060110336}, "loss": 1.0024921894073486, "ce_loss": 1.0014845132827759, "perplexity": 2.722320150669571, "z_loss": 10.076813340187073, "grad_norm": 0.14151111245155334, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1087264.3558919267, "data_wait_seconds": 2.0563602447509766e-05, "eta_seconds": 23691.119772678703}
{"step": 8100, "total_steps": 57221, "tokens": 4246732800, "source": "stack_v3", "source_tokens": {"stack_v3": 3185049600, "fineweb_edu": 1061683200}, "loss": 1.0881335139274597, "ce_loss": 1.0869677662849426, "perplexity": 2.965269038026769, "z_loss": 11.657381415367126, "grad_norm": 0.1711674928665161, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1169349.1960168527, "data_wait_seconds": 2.1457672119140625e-05, "eta_seconds": 22023.589948770823}
{"step": 8110, "total_steps": 57221, "tokens": 4251975680, "source": "stack_v3", "source_tokens": {"stack_v3": 3189243904, "fineweb_edu": 1062731776}, "loss": 1.1278281062841415, "ce_loss": 1.1267794072628021, "perplexity": 3.0857026884738956, "z_loss": 10.48717749118805, "grad_norm": 0.13533884286880493, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144966.0351432648, "data_wait_seconds": 1.7881393432617188e-05, "eta_seconds": 22488.024561163737}
{"step": 8120, "total_steps": 57221, "tokens": 4257218560, "source": "stack_v3", "source_tokens": {"stack_v3": 3192913920, "fineweb_edu": 1064304640}, "loss": 1.0344264954328537, "ce_loss": 1.033541351556778, "perplexity": 2.811002978628207, "z_loss": 8.851487874984741, "grad_norm": 0.14060477912425995, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128551.3995336602, "data_wait_seconds": 1.806020736694336e-05, "eta_seconds": 22810.464326779824}
{"step": 8130, "total_steps": 57221, "tokens": 4262461440, "source": "stack_v3", "source_tokens": {"stack_v3": 3197108224, "fineweb_edu": 1065353216}, "loss": 1.096185028553009, "ce_loss": 1.0953951478004456, "perplexity": 2.9903640857547362, "z_loss": 7.899018287658691, "grad_norm": 0.17087280750274658, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1161944.9027046557, "data_wait_seconds": 1.800060272216797e-05, "eta_seconds": 22150.394997293595}
{"step": 8140, "total_steps": 57221, "tokens": 4267704320, "source": "stack_v3", "source_tokens": {"stack_v3": 3200778240, "fineweb_edu": 1066926080}, "loss": 1.0650053173303604, "ce_loss": 1.0640080720186234, "perplexity": 2.8979629873285058, "z_loss": 9.972739934921265, "grad_norm": 0.11890626698732376, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120937.621807394, "data_wait_seconds": 2.0742416381835938e-05, "eta_seconds": 22956.046063035494}
{"step": 8150, "total_steps": 57221, "tokens": 4272947200, "source": "stack_v3", "source_tokens": {"stack_v3": 3204972544, "fineweb_edu": 1067974656}, "loss": 1.0041493475437164, "ce_loss": 1.0032837092876434, "perplexity": 2.7272225470721723, "z_loss": 8.65631115436554, "grad_norm": 0.10647881031036377, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1170038.6313789443, "data_wait_seconds": 1.8477439880371094e-05, "eta_seconds": 21988.208004448097}
{"step": 8160, "total_steps": 57221, "tokens": 4278190080, "source": "stack_v3", "source_tokens": {"stack_v3": 3208642560, "fineweb_edu": 1069547520}, "loss": 0.8617256805300713, "ce_loss": 0.8609744384884834, "perplexity": 2.3654645707511457, "z_loss": 7.5125333070755005, "grad_norm": 0.1454622447490692, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1126214.915461282, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 22839.166456488194}
{"step": 8170, "total_steps": 57221, "tokens": 4283432960, "source": "stack_v3", "source_tokens": {"stack_v3": 3212836864, "fineweb_edu": 1070596096}, "loss": 1.1789704710245132, "ce_loss": 1.1779737025499344, "perplexity": 3.2477865492676745, "z_loss": 9.96762752532959, "grad_norm": 0.11085480451583862, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1082209.3192896629, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 23763.024935767287}
{"step": 8180, "total_steps": 57221, "tokens": 4288675840, "source": "stack_v3", "source_tokens": {"stack_v3": 3216506880, "fineweb_edu": 1072168960}, "loss": 1.1272452622652054, "ce_loss": 1.126253992319107, "perplexity": 3.08408184001586, "z_loss": 9.912787675857544, "grad_norm": 0.17688030004501343, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1138120.0161788338, "data_wait_seconds": 2.0623207092285156e-05, "eta_seconds": 22591.048215041636}
{"step": 8190, "total_steps": 57221, "tokens": 4293918720, "source": "stack_v3", "source_tokens": {"stack_v3": 3220701184, "fineweb_edu": 1073217536}, "loss": 1.1359029114246368, "ce_loss": 1.1348699778318405, "perplexity": 3.1107690476728016, "z_loss": 10.329623460769653, "grad_norm": 0.11193709820508957, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1107546.657265127, "data_wait_seconds": 1.919269561767578e-05, "eta_seconds": 23209.93080641516}
{"step": 8200, "total_steps": 57221, "tokens": 4299161600, "source": "stack_v3", "source_tokens": {"stack_v3": 3224371200, "fineweb_edu": 1074790400}, "loss": 1.2390059530735016, "ce_loss": 1.23810613155365, "perplexity": 3.449075180738142, "z_loss": 8.998227596282959, "grad_norm": 0.17666871845722198, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1143140.558874937, "data_wait_seconds": 2.6702880859375e-05, "eta_seconds": 22482.658147738548}
{"step": 8210, "total_steps": 57221, "tokens": 4304404480, "source": "stack_v3", "source_tokens": {"stack_v3": 3228565504, "fineweb_edu": 1075838976}, "loss": 1.100339263677597, "ce_loss": 1.099409207701683, "perplexity": 3.002391709973743, "z_loss": 9.300708293914795, "grad_norm": 0.09380647540092468, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1155540.250105871, "data_wait_seconds": 1.7940998077392578e-05, "eta_seconds": 22236.867575703887}
{"step": 8220, "total_steps": 57221, "tokens": 4309647360, "source": "stack_v3", "source_tokens": {"stack_v3": 3232235520, "fineweb_edu": 1077411840}, "loss": 1.0697524845600128, "ce_loss": 1.0688684731721878, "perplexity": 2.912082535513163, "z_loss": 8.840253233909607, "grad_norm": 0.1481267511844635, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1092958.8476528397, "data_wait_seconds": 2.5272369384765625e-05, "eta_seconds": 23505.32473859447}
{"step": 8230, "total_steps": 57221, "tokens": 4314890240, "source": "stack_v3", "source_tokens": {"stack_v3": 3236429824, "fineweb_edu": 1078460416}, "loss": 1.202857032418251, "ce_loss": 1.2020553052425385, "perplexity": 3.326947793811629, "z_loss": 8.017432928085327, "grad_norm": 0.18565566837787628, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1153883.7856391, "data_wait_seconds": 1.895427703857422e-05, "eta_seconds": 22259.702475820668}
{"step": 8240, "total_steps": 57221, "tokens": 4320133120, "source": "stack_v3", "source_tokens": {"stack_v3": 3240099840, "fineweb_edu": 1080033280}, "loss": 0.9717807918787003, "ce_loss": 0.9708992540836334, "perplexity": 2.6403177085987335, "z_loss": 8.815406203269958, "grad_norm": 0.12642516195774078, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1102300.724183365, "data_wait_seconds": 1.9252300262451172e-05, "eta_seconds": 23296.607102408307}
{"step": 8250, "total_steps": 57221, "tokens": 4325376000, "source": "stack_v3", "source_tokens": {"stack_v3": 3244294144, "fineweb_edu": 1081081856}, "loss": 1.0282579511404037, "ce_loss": 1.0273850560188293, "perplexity": 2.7937507728605575, "z_loss": 8.728748798370361, "grad_norm": 0.1049858033657074, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128005.0928209107, "data_wait_seconds": 1.811981201171875e-05, "eta_seconds": 22761.08872504556}
{"step": 8260, "total_steps": 57221, "tokens": 4330618880, "source": "stack_v3", "source_tokens": {"stack_v3": 3247964160, "fineweb_edu": 1082654720}, "loss": 0.9946418702602386, "ce_loss": 0.993743047118187, "perplexity": 2.7013267659675386, "z_loss": 8.988163471221924, "grad_norm": 0.1345415860414505, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119160.5049289363, "data_wait_seconds": 1.9073486328125e-05, "eta_seconds": 22936.28215698153}
{"step": 8270, "total_steps": 57221, "tokens": 4335861760, "source": "stack_v3", "source_tokens": {"stack_v3": 3252158464, "fineweb_edu": 1083703296}, "loss": 1.103891208767891, "ce_loss": 1.1030108481645584, "perplexity": 3.0132247420748244, "z_loss": 8.803662300109863, "grad_norm": 0.12764397263526917, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1108534.6564657688, "data_wait_seconds": 2.0205974578857422e-05, "eta_seconds": 23151.408113682643}
{"step": 8280, "total_steps": 57221, "tokens": 4341104640, "source": "stack_v3", "source_tokens": {"stack_v3": 3255828480, "fineweb_edu": 1085276160}, "loss": 1.1229104846715927, "ce_loss": 1.122085988521576, "perplexity": 3.0712541267441127, "z_loss": 8.245056688785553, "grad_norm": 0.15205511450767517, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1106307.9541151177, "data_wait_seconds": 1.9609928131103516e-05, "eta_seconds": 23193.266634807223}
{"step": 8290, "total_steps": 57221, "tokens": 4346347520, "source": "stack_v3", "source_tokens": {"stack_v3": 3260022784, "fineweb_edu": 1086324736}, "loss": 0.9293644726276398, "ce_loss": 0.9285020232200623, "perplexity": 2.530715383938083, "z_loss": 8.624661445617676, "grad_norm": 0.09754551202058792, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125429.4233230627, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 22794.54575148062}
{"step": 8300, "total_steps": 57221, "tokens": 4351590400, "source": "stack_v3", "source_tokens": {"stack_v3": 3263692800, "fineweb_edu": 1087897600}, "loss": 0.9684832096099854, "ce_loss": 0.967500127851963, "perplexity": 2.6313581713411334, "z_loss": 9.830782651901245, "grad_norm": 0.13023792207241058, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1078676.6114882897, "data_wait_seconds": 1.6748905181884766e-05, "eta_seconds": 23777.663598927902}
{"step": 8310, "total_steps": 57221, "tokens": 4356833280, "source": "stack_v3", "source_tokens": {"stack_v3": 3267887104, "fineweb_edu": 1088946176}, "loss": 0.9574625343084335, "ce_loss": 0.9563778936862946, "perplexity": 2.6022537427482253, "z_loss": 10.846208691596985, "grad_norm": 0.09901244938373566, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1166649.816413491, "data_wait_seconds": 2.0503997802734375e-05, "eta_seconds": 21980.174649863737}
{"step": 8320, "total_steps": 57221, "tokens": 4362076160, "source": "stack_v3", "source_tokens": {"stack_v3": 3271557120, "fineweb_edu": 1090519040}, "loss": 0.9109458476305008, "ce_loss": 0.910119466483593, "perplexity": 2.4846193443911213, "z_loss": 8.26374638080597, "grad_norm": 0.1524169147014618, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1095301.398729424, "data_wait_seconds": 1.8894672393798828e-05, "eta_seconds": 23407.186250050087}
{"step": 8330, "total_steps": 57221, "tokens": 4367319040, "source": "stack_v3", "source_tokens": {"stack_v3": 3275751424, "fineweb_edu": 1091567616}, "loss": 0.9517684578895569, "ce_loss": 0.9509033560752869, "perplexity": 2.5880465312003067, "z_loss": 8.651267170906067, "grad_norm": 0.11281195282936096, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1171240.379739918, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 21885.072785564236}
{"step": 8340, "total_steps": 57221, "tokens": 4372561920, "source": "stack_v3", "source_tokens": {"stack_v3": 3279421440, "fineweb_edu": 1093140480}, "loss": 1.2078661769628525, "ce_loss": 1.2069299966096878, "perplexity": 3.3432052303004753, "z_loss": 9.361786842346191, "grad_norm": 0.14975762367248535, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1125806.2973999628, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 22763.630066012498}
{"step": 8350, "total_steps": 57221, "tokens": 4377804800, "source": "stack_v3", "source_tokens": {"stack_v3": 3283615744, "fineweb_edu": 1094189056}, "loss": 1.0172201246023178, "ce_loss": 1.0163523107767105, "perplexity": 2.7630974384583613, "z_loss": 8.678281545639038, "grad_norm": 0.12911374866962433, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1111641.4704267825, "data_wait_seconds": 2.008676528930664e-05, "eta_seconds": 23048.973865794247}
{"step": 8360, "total_steps": 57221, "tokens": 4383047680, "source": "stack_v3", "source_tokens": {"stack_v3": 3287285760, "fineweb_edu": 1095761920}, "loss": 0.9842367619276047, "ce_loss": 0.9834931492805481, "perplexity": 2.6737798602486413, "z_loss": 7.436078429222107, "grad_norm": 0.15334174036979675, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1146160.5334098614, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 22350.230681725545}
{"step": 8370, "total_steps": 57221, "tokens": 4388290560, "source": "stack_v3", "source_tokens": {"stack_v3": 3291480064, "fineweb_edu": 1096810496}, "loss": 0.9570851475000381, "ce_loss": 0.9561402797698975, "perplexity": 2.6016354845012426, "z_loss": 9.44867491722107, "grad_norm": 0.09467516839504242, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1081417.3526317654, "data_wait_seconds": 1.9431114196777344e-05, "eta_seconds": 23683.4644623287}
{"step": 8380, "total_steps": 57221, "tokens": 4393533440, "source": "stack_v3", "source_tokens": {"stack_v3": 3295150080, "fineweb_edu": 1098383360}, "loss": 1.056006595492363, "ce_loss": 1.054952621459961, "perplexity": 2.8718390871348736, "z_loss": 10.53971016407013, "grad_norm": 0.1126457080245018, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1144408.9063277128, "data_wait_seconds": 1.621246337890625e-05, "eta_seconds": 22375.27724436228}
{"step": 8390, "total_steps": 57221, "tokens": 4398776320, "source": "stack_v3", "source_tokens": {"stack_v3": 3299344384, "fineweb_edu": 1099431936}, "loss": 1.0459486246109009, "ce_loss": 1.0451417416334152, "perplexity": 2.8438015801672316, "z_loss": 8.068701922893524, "grad_norm": 0.11816643178462982, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1157178.0937463264, "data_wait_seconds": 1.7523765563964844e-05, "eta_seconds": 22123.840589754745}
{"step": 8400, "total_steps": 57221, "tokens": 4404019200, "source": "stack_v3", "source_tokens": {"stack_v3": 3303014400, "fineweb_edu": 1101004800}, "loss": 1.1113128960132599, "ce_loss": 1.1104168444871902, "perplexity": 3.0356235136652248, "z_loss": 8.960373163223267, "grad_norm": 0.1498536318540573, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1110228.0485027705, "data_wait_seconds": 2.1636486053466797e-05, "eta_seconds": 23054.705593610415}
{"step": 8410, "total_steps": 57221, "tokens": 4409262080, "source": "stack_v3", "source_tokens": {"stack_v3": 3307208704, "fineweb_edu": 1102053376}, "loss": 0.8578480780124664, "ce_loss": 0.8571553826332092, "perplexity": 2.3564479578649737, "z_loss": 6.9270113706588745, "grad_norm": 0.09802365303039551, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1158190.4187917488, "data_wait_seconds": 1.621246337890625e-05, "eta_seconds": 22095.449508809488}
{"step": 8420, "total_steps": 57221, "tokens": 4414504960, "source": "stack_v3", "source_tokens": {"stack_v3": 3310878720, "fineweb_edu": 1103626240}, "loss": 1.0318223983049393, "ce_loss": 1.0307997912168503, "perplexity": 2.803306998663652, "z_loss": 10.226067185401917, "grad_norm": 0.1973116397857666, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1134973.2425027138, "data_wait_seconds": 1.6033649444580078e-05, "eta_seconds": 22542.81782324822}
{"step": 8430, "total_steps": 57221, "tokens": 4419747840, "source": "stack_v3", "source_tokens": {"stack_v3": 3315073024, "fineweb_edu": 1104674816}, "loss": 1.0371315628290176, "ce_loss": 1.0361830741167068, "perplexity": 2.8184386858263077, "z_loss": 9.485010981559753, "grad_norm": 0.10146903991699219, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1135964.3840715406, "data_wait_seconds": 1.8775463104248047e-05, "eta_seconds": 22518.533607818652}
{"step": 8440, "total_steps": 57221, "tokens": 4424990720, "source": "stack_v3", "source_tokens": {"stack_v3": 3318743040, "fineweb_edu": 1106247680}, "loss": 1.0749697387218475, "ce_loss": 1.0740499943494797, "perplexity": 2.927210712478643, "z_loss": 9.197495102882385, "grad_norm": 0.1810886412858963, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1159612.7933795932, "data_wait_seconds": 2.1398067474365234e-05, "eta_seconds": 22054.783653657185}
{"step": 8450, "total_steps": 57221, "tokens": 4430233600, "source": "stack_v3", "source_tokens": {"stack_v3": 3322937344, "fineweb_edu": 1107296256}, "loss": 1.0086278170347214, "ce_loss": 1.0077358037233353, "perplexity": 2.739391467891063, "z_loss": 8.920219421386719, "grad_norm": 0.1001851037144661, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131325.6359855495, "data_wait_seconds": 1.728534698486328e-05, "eta_seconds": 22601.597264898013}
{"step": 8460, "total_steps": 57221, "tokens": 4435476480, "source": "stack_v3", "source_tokens": {"stack_v3": 3326607360, "fineweb_edu": 1108869120}, "loss": 1.070887342095375, "ce_loss": 1.0698541849851608, "perplexity": 2.9149544248638657, "z_loss": 10.331451296806335, "grad_norm": 0.14863556623458862, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1131085.5413367532, "data_wait_seconds": 1.7464160919189453e-05, "eta_seconds": 22601.759624463968}
{"step": 8470, "total_steps": 57221, "tokens": 4440719360, "source": "stack_v3", "source_tokens": {"stack_v3": 3330801664, "fineweb_edu": 1109917696}, "loss": 1.034514233469963, "ce_loss": 1.0335809886455536, "perplexity": 2.8111144008110305, "z_loss": 9.332366943359375, "grad_norm": 0.10669173300266266, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1156283.0439231475, "data_wait_seconds": 1.6391277313232422e-05, "eta_seconds": 22104.692077192478}
{"step": 8480, "total_steps": 57221, "tokens": 4445962240, "source": "stack_v3", "source_tokens": {"stack_v3": 3334471680, "fineweb_edu": 1111490560}, "loss": 1.1310612708330154, "ce_loss": 1.1301077008247375, "perplexity": 3.095989922837507, "z_loss": 9.535940170288086, "grad_norm": 0.16119413077831268, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1099656.0676004793, "data_wait_seconds": 2.110004425048828e-05, "eta_seconds": 23238.209211868}
{"step": 8490, "total_steps": 57221, "tokens": 4451205120, "source": "stack_v3", "source_tokens": {"stack_v3": 3338665984, "fineweb_edu": 1112539136}, "loss": 1.0504826307296753, "ce_loss": 1.049654945731163, "perplexity": 2.8566652434459336, "z_loss": 8.27699077129364, "grad_norm": 0.12376049906015396, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1137844.1168762075, "data_wait_seconds": 1.913309097290039e-05, "eta_seconds": 22453.68631877331}
{"step": 8500, "total_steps": 57221, "tokens": 4456448000, "source": "stack_v3", "source_tokens": {"stack_v3": 3342336000, "fineweb_edu": 1114112000}, "loss": 1.1506102830171585, "ce_loss": 1.1496307253837585, "perplexity": 3.1570268845200244, "z_loss": 9.79543673992157, "grad_norm": 0.16796956956386566, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1128636.429428327, "data_wait_seconds": 1.6868114471435547e-05, "eta_seconds": 22632.223569939375}
{"step": 8510, "total_steps": 57221, "tokens": 4461690880, "source": "stack_v3", "source_tokens": {"stack_v3": 3346530304, "fineweb_edu": 1115160576}, "loss": 1.0502197742462158, "ce_loss": 1.049338936805725, "perplexity": 2.8557626543526324, "z_loss": 8.808321595191956, "grad_norm": 0.18286681175231934, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1133325.294600646, "data_wait_seconds": 1.990795135498047e-05, "eta_seconds": 22533.96199808549}
{"step": 8520, "total_steps": 57221, "tokens": 4466933760, "source": "stack_v3", "source_tokens": {"stack_v3": 3350200320, "fineweb_edu": 1116733440}, "loss": 1.0852145552635193, "ce_loss": 1.0842850059270859, "perplexity": 2.9573245930869603, "z_loss": 9.295403718948364, "grad_norm": 0.16677145659923553, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1119520.595474802, "data_wait_seconds": 1.9729137420654297e-05, "eta_seconds": 22807.142935294658}
{"step": 8530, "total_steps": 57221, "tokens": 4472176640, "source": "stack_v3", "source_tokens": {"stack_v3": 3354394624, "fineweb_edu": 1117782016}, "loss": 1.042045772075653, "ce_loss": 1.041115716099739, "perplexity": 2.8323753789155903, "z_loss": 9.300432801246643, "grad_norm": 0.14744263887405396, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1089259.3553346866, "data_wait_seconds": 2.014636993408203e-05, "eta_seconds": 23435.945934250252}
{"step": 8540, "total_steps": 57221, "tokens": 4477419520, "source": "stack_v3", "source_tokens": {"stack_v3": 3358064640, "fineweb_edu": 1119354880}, "loss": 0.9391958564519882, "ce_loss": 0.9382954090833664, "perplexity": 2.5556214143241616, "z_loss": 9.004479169845581, "grad_norm": 0.13936235010623932, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1186684.331411364, "data_wait_seconds": 1.615285873413086e-05, "eta_seconds": 21507.472378644394}
{"step": 8550, "total_steps": 57221, "tokens": 4482662400, "source": "stack_v3", "source_tokens": {"stack_v3": 3362258944, "fineweb_edu": 1120403456}, "loss": 1.1368563026189804, "ce_loss": 1.1358834654092789, "perplexity": 3.113923371622968, "z_loss": 9.728413939476013, "grad_norm": 0.11126111447811127, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1107120.5908165344, "data_wait_seconds": 1.722574234008789e-05, "eta_seconds": 23048.381370253628}
{"step": 8560, "total_steps": 57221, "tokens": 4487905280, "source": "stack_v3", "source_tokens": {"stack_v3": 3365928960, "fineweb_edu": 1121976320}, "loss": 0.8119216486811638, "ce_loss": 0.8112209886312485, "perplexity": 2.2506543330599613, "z_loss": 7.006622314453125, "grad_norm": 0.183358296751976, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1053066.8514400334, "data_wait_seconds": 2.467632293701172e-05, "eta_seconds": 24226.472122936044}
{"step": 8570, "total_steps": 57221, "tokens": 4493148160, "source": "stack_v3", "source_tokens": {"stack_v3": 3370123264, "fineweb_edu": 1123024896}, "loss": 0.8832052201032639, "ce_loss": 0.8823923915624619, "perplexity": 2.416674427475433, "z_loss": 8.128270268440247, "grad_norm": 0.09105930477380753, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1133878.346582571, "data_wait_seconds": 1.9311904907226562e-05, "eta_seconds": 22495.227920063775}
{"step": 8580, "total_steps": 57221, "tokens": 4498391040, "source": "stack_v3", "source_tokens": {"stack_v3": 3373793280, "fineweb_edu": 1124597760}, "loss": 0.7366374358534813, "ce_loss": 0.7359168753027916, "perplexity": 2.0873949962973226, "z_loss": 7.205585420131683, "grad_norm": 0.1559639573097229, "muon_lr": 0.02, "adamw_lr": 0.002, "ngram_lr": 0.001, "weight_decay": 0.1, "tokens_per_second": 1120614.7648744239, "data_wait_seconds": 1.722574234008789e-05, "eta_seconds": 22756.802568862913}