silly_nobel / checkpoint-1600 /trainer_state.json
tomekkorbak's picture
Training in progress, step 1600
64c14be
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.6339144215530903,
"global_step": 1600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 3.846153846153847e-06,
"loss": 2.7825,
"theoretical_loss": 3.3518567762894107,
"tokens_seen": 2969305088
},
{
"epoch": 0.0,
"learning_rate": 3.846153846153846e-05,
"loss": 2.4663,
"theoretical_loss": 3.3517550749686795,
"tokens_seen": 2970484736
},
{
"epoch": 0.01,
"learning_rate": 7.692307692307693e-05,
"loss": 1.9063,
"theoretical_loss": 3.351642134115535,
"tokens_seen": 2971795456
},
{
"epoch": 0.01,
"learning_rate": 9.983987189751803e-05,
"loss": 1.4439,
"theoretical_loss": 3.351529257004948,
"tokens_seen": 2973106176
},
{
"epoch": 0.02,
"learning_rate": 9.943955164131305e-05,
"loss": 1.2731,
"theoretical_loss": 3.3514164435728655,
"tokens_seen": 2974416896
},
{
"epoch": 0.02,
"learning_rate": 9.90392313851081e-05,
"loss": 1.2079,
"theoretical_loss": 3.3513036937553267,
"tokens_seen": 2975727616
},
{
"epoch": 0.02,
"learning_rate": 9.863891112890312e-05,
"loss": 1.2085,
"theoretical_loss": 3.3511910074884628,
"tokens_seen": 2977038336
},
{
"epoch": 0.03,
"learning_rate": 9.823859087269817e-05,
"loss": 1.2031,
"theoretical_loss": 3.3510783847084977,
"tokens_seen": 2978349056
},
{
"epoch": 0.03,
"learning_rate": 9.78382706164932e-05,
"loss": 1.2176,
"theoretical_loss": 3.350965825351748,
"tokens_seen": 2979659776
},
{
"epoch": 0.04,
"learning_rate": 9.743795036028824e-05,
"loss": 1.1891,
"theoretical_loss": 3.3508533293546208,
"tokens_seen": 2980970496
},
{
"epoch": 0.04,
"learning_rate": 9.703763010408326e-05,
"loss": 1.2086,
"theoretical_loss": 3.3507408966536154,
"tokens_seen": 2982281216
},
{
"epoch": 0.04,
"learning_rate": 9.663730984787832e-05,
"loss": 1.18,
"theoretical_loss": 3.3506285271853233,
"tokens_seen": 2983591936
},
{
"epoch": 0.05,
"learning_rate": 9.623698959167334e-05,
"loss": 1.1709,
"theoretical_loss": 3.350516220886427,
"tokens_seen": 2984902656
},
{
"epoch": 0.05,
"objective/train/docs_used": 1000073,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.024436354637146,
"objective/train/theoretical_loss": 3.3504600914057403,
"objective/train/tokens_used": 36842976,
"theoretical_loss": 3.3504600914057403,
"tokens_seen": 2985558016
},
{
"epoch": 0.05,
"learning_rate": 9.583666933546838e-05,
"loss": 1.1659,
"theoretical_loss": 3.3504039776936994,
"tokens_seen": 2986213376
},
{
"epoch": 0.06,
"learning_rate": 9.543634907926342e-05,
"loss": 1.1826,
"theoretical_loss": 3.350291797544005,
"tokens_seen": 2987524096
},
{
"epoch": 0.06,
"learning_rate": 9.503602882305846e-05,
"loss": 1.1854,
"theoretical_loss": 3.3501796803742994,
"tokens_seen": 2988834816
},
{
"epoch": 0.06,
"learning_rate": 9.463570856685348e-05,
"loss": 1.1901,
"theoretical_loss": 3.3500676261216285,
"tokens_seen": 2990145536
},
{
"epoch": 0.07,
"learning_rate": 9.423538831064852e-05,
"loss": 1.1854,
"theoretical_loss": 3.3499556347231287,
"tokens_seen": 2991456256
},
{
"epoch": 0.07,
"learning_rate": 9.383506805444356e-05,
"loss": 1.2042,
"theoretical_loss": 3.349843706116027,
"tokens_seen": 2992766976
},
{
"epoch": 0.08,
"learning_rate": 9.34347477982386e-05,
"loss": 1.1804,
"theoretical_loss": 3.3497318402376397,
"tokens_seen": 2994077696
},
{
"epoch": 0.08,
"learning_rate": 9.303442754203363e-05,
"loss": 1.1869,
"theoretical_loss": 3.349620037025374,
"tokens_seen": 2995388416
},
{
"epoch": 0.08,
"learning_rate": 9.263410728582867e-05,
"loss": 1.1783,
"theoretical_loss": 3.349508296416727,
"tokens_seen": 2996699136
},
{
"epoch": 0.09,
"learning_rate": 9.22337870296237e-05,
"loss": 1.1659,
"theoretical_loss": 3.3493966183492847,
"tokens_seen": 2998009856
},
{
"epoch": 0.09,
"learning_rate": 9.183346677341874e-05,
"loss": 1.1763,
"theoretical_loss": 3.349285002760723,
"tokens_seen": 2999320576
},
{
"epoch": 0.1,
"learning_rate": 9.143314651721377e-05,
"loss": 1.151,
"theoretical_loss": 3.3491734495888066,
"tokens_seen": 3000631296
},
{
"epoch": 0.1,
"objective/train/docs_used": 1005438,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.0497773885726929,
"objective/train/theoretical_loss": 3.3490619587713906,
"objective/train/tokens_used": 53226976,
"theoretical_loss": 3.3490619587713906,
"tokens_seen": 3001942016
},
{
"epoch": 0.1,
"learning_rate": 9.103282626100881e-05,
"loss": 1.1792,
"theoretical_loss": 3.3490619587713906,
"tokens_seen": 3001942016
},
{
"epoch": 0.1,
"learning_rate": 9.063250600480385e-05,
"loss": 1.1536,
"theoretical_loss": 3.348950530246418,
"tokens_seen": 3003252736
},
{
"epoch": 0.11,
"learning_rate": 9.023218574859889e-05,
"loss": 1.1929,
"theoretical_loss": 3.348839163951921,
"tokens_seen": 3004563456
},
{
"epoch": 0.11,
"learning_rate": 8.983186549239391e-05,
"loss": 1.1682,
"theoretical_loss": 3.3487278598260195,
"tokens_seen": 3005874176
},
{
"epoch": 0.11,
"learning_rate": 8.943154523618895e-05,
"loss": 1.148,
"theoretical_loss": 3.348616617806923,
"tokens_seen": 3007184896
},
{
"epoch": 0.12,
"learning_rate": 8.903122497998399e-05,
"loss": 1.1768,
"theoretical_loss": 3.34850543783293,
"tokens_seen": 3008495616
},
{
"epoch": 0.12,
"learning_rate": 8.863090472377903e-05,
"loss": 1.1853,
"theoretical_loss": 3.3483943198424253,
"tokens_seen": 3009806336
},
{
"epoch": 0.13,
"learning_rate": 8.823058446757406e-05,
"loss": 1.1791,
"theoretical_loss": 3.348283263773882,
"tokens_seen": 3011117056
},
{
"epoch": 0.13,
"learning_rate": 8.78302642113691e-05,
"loss": 1.18,
"theoretical_loss": 3.3481722695658633,
"tokens_seen": 3012427776
},
{
"epoch": 0.13,
"learning_rate": 8.742994395516414e-05,
"loss": 1.1761,
"theoretical_loss": 3.3480613371570165,
"tokens_seen": 3013738496
},
{
"epoch": 0.14,
"learning_rate": 8.702962369895917e-05,
"loss": 1.1834,
"theoretical_loss": 3.347950466486079,
"tokens_seen": 3015049216
},
{
"epoch": 0.14,
"learning_rate": 8.66293034427542e-05,
"loss": 1.1986,
"theoretical_loss": 3.3478396574918747,
"tokens_seen": 3016359936
},
{
"epoch": 0.15,
"learning_rate": 8.622898318654924e-05,
"loss": 1.1895,
"theoretical_loss": 3.347728910113315,
"tokens_seen": 3017670656
},
{
"epoch": 0.15,
"objective/train/docs_used": 1011213,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.2239086627960205,
"objective/train/theoretical_loss": 3.3476735595108353,
"objective/train/tokens_used": 69610976,
"theoretical_loss": 3.3476735595108353,
"tokens_seen": 3018326016
},
{
"epoch": 0.15,
"learning_rate": 8.582866293034428e-05,
"loss": 1.1745,
"theoretical_loss": 3.3476182242893975,
"tokens_seen": 3018981376
},
{
"epoch": 0.15,
"learning_rate": 8.542834267413932e-05,
"loss": 1.1915,
"theoretical_loss": 3.3475075999592074,
"tokens_seen": 3020292096
},
{
"epoch": 0.16,
"learning_rate": 8.502802241793434e-05,
"loss": 1.1925,
"theoretical_loss": 3.347397037061916,
"tokens_seen": 3021602816
},
{
"epoch": 0.16,
"learning_rate": 8.462770216172938e-05,
"loss": 1.1872,
"theoretical_loss": 3.3472865355367825,
"tokens_seen": 3022913536
},
{
"epoch": 0.17,
"learning_rate": 8.422738190552442e-05,
"loss": 1.1753,
"theoretical_loss": 3.3471760953231504,
"tokens_seen": 3024224256
},
{
"epoch": 0.17,
"learning_rate": 8.382706164931946e-05,
"loss": 1.2088,
"theoretical_loss": 3.3470657163604507,
"tokens_seen": 3025534976
},
{
"epoch": 0.17,
"learning_rate": 8.34267413931145e-05,
"loss": 1.1959,
"theoretical_loss": 3.3469553985882,
"tokens_seen": 3026845696
},
{
"epoch": 0.18,
"learning_rate": 8.302642113690953e-05,
"loss": 1.2173,
"theoretical_loss": 3.3468451419460017,
"tokens_seen": 3028156416
},
{
"epoch": 0.18,
"learning_rate": 8.262610088070457e-05,
"loss": 1.1876,
"theoretical_loss": 3.3467349463735436,
"tokens_seen": 3029467136
},
{
"epoch": 0.19,
"learning_rate": 8.22257806244996e-05,
"loss": 1.1839,
"theoretical_loss": 3.3466248118105995,
"tokens_seen": 3030777856
},
{
"epoch": 0.19,
"learning_rate": 8.182546036829464e-05,
"loss": 1.198,
"theoretical_loss": 3.3465147381970293,
"tokens_seen": 3032088576
},
{
"epoch": 0.19,
"learning_rate": 8.142514011208967e-05,
"loss": 1.1891,
"theoretical_loss": 3.346404725472777,
"tokens_seen": 3033399296
},
{
"epoch": 0.2,
"objective/train/docs_used": 1016482,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.2322360277175903,
"objective/train/theoretical_loss": 3.3462947735778727,
"objective/train/tokens_used": 85994976,
"theoretical_loss": 3.3462947735778727,
"tokens_seen": 3034710016
},
{
"epoch": 0.2,
"learning_rate": 8.102481985588471e-05,
"loss": 1.1985,
"theoretical_loss": 3.3462947735778727,
"tokens_seen": 3034710016
},
{
"epoch": 0.2,
"learning_rate": 8.062449959967975e-05,
"loss": 1.1688,
"theoretical_loss": 3.346184882452431,
"tokens_seen": 3036020736
},
{
"epoch": 0.21,
"learning_rate": 8.022417934347479e-05,
"loss": 1.1874,
"theoretical_loss": 3.3460750520366513,
"tokens_seen": 3037331456
},
{
"epoch": 0.21,
"learning_rate": 7.982385908726981e-05,
"loss": 1.1835,
"theoretical_loss": 3.3459652822708175,
"tokens_seen": 3038642176
},
{
"epoch": 0.21,
"learning_rate": 7.942353883106487e-05,
"loss": 1.1842,
"theoretical_loss": 3.345855573095298,
"tokens_seen": 3039952896
},
{
"epoch": 0.22,
"learning_rate": 7.902321857485989e-05,
"loss": 1.1882,
"theoretical_loss": 3.3457459244505463,
"tokens_seen": 3041263616
},
{
"epoch": 0.22,
"learning_rate": 7.862289831865493e-05,
"loss": 1.2044,
"theoretical_loss": 3.3456363362770984,
"tokens_seen": 3042574336
},
{
"epoch": 0.23,
"learning_rate": 7.822257806244996e-05,
"loss": 1.168,
"theoretical_loss": 3.345526808515576,
"tokens_seen": 3043885056
},
{
"epoch": 0.23,
"learning_rate": 7.782225780624501e-05,
"loss": 1.1908,
"theoretical_loss": 3.3454173411066828,
"tokens_seen": 3045195776
},
{
"epoch": 0.23,
"learning_rate": 7.742193755004003e-05,
"loss": 1.1942,
"theoretical_loss": 3.345307933991209,
"tokens_seen": 3046506496
},
{
"epoch": 0.24,
"learning_rate": 7.702161729383507e-05,
"loss": 1.1965,
"theoretical_loss": 3.3451985871100254,
"tokens_seen": 3047817216
},
{
"epoch": 0.24,
"learning_rate": 7.662129703763011e-05,
"loss": 1.1727,
"theoretical_loss": 3.345089300404087,
"tokens_seen": 3049127936
},
{
"epoch": 0.25,
"learning_rate": 7.622097678142515e-05,
"loss": 1.1993,
"theoretical_loss": 3.344980073814434,
"tokens_seen": 3050438656
},
{
"epoch": 0.25,
"objective/train/docs_used": 1022217,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 0.9348351955413818,
"objective/train/theoretical_loss": 3.3449254830448116,
"objective/train/tokens_used": 102378976,
"theoretical_loss": 3.3449254830448116,
"tokens_seen": 3051094016
},
{
"epoch": 0.25,
"learning_rate": 7.582065652522018e-05,
"loss": 1.1904,
"theoretical_loss": 3.3448709072821874,
"tokens_seen": 3051749376
},
{
"epoch": 0.25,
"learning_rate": 7.542033626901522e-05,
"loss": 1.1939,
"theoretical_loss": 3.344761800748551,
"tokens_seen": 3053060096
},
{
"epoch": 0.26,
"learning_rate": 7.502001601281026e-05,
"loss": 1.1901,
"theoretical_loss": 3.3446527541548132,
"tokens_seen": 3054370816
},
{
"epoch": 0.26,
"learning_rate": 7.46196957566053e-05,
"loss": 1.2002,
"theoretical_loss": 3.3445437674423437,
"tokens_seen": 3055681536
},
{
"epoch": 0.27,
"learning_rate": 7.421937550040032e-05,
"loss": 1.2222,
"theoretical_loss": 3.3444348405525948,
"tokens_seen": 3056992256
},
{
"epoch": 0.27,
"learning_rate": 7.381905524419536e-05,
"loss": 1.2025,
"theoretical_loss": 3.3443259734271016,
"tokens_seen": 3058302976
},
{
"epoch": 0.27,
"learning_rate": 7.34187349879904e-05,
"loss": 1.1916,
"theoretical_loss": 3.344217166007481,
"tokens_seen": 3059613696
},
{
"epoch": 0.28,
"learning_rate": 7.301841473178544e-05,
"loss": 1.1965,
"theoretical_loss": 3.344108418235432,
"tokens_seen": 3060924416
},
{
"epoch": 0.28,
"learning_rate": 7.261809447558046e-05,
"loss": 1.1738,
"theoretical_loss": 3.343999730052735,
"tokens_seen": 3062235136
},
{
"epoch": 0.29,
"learning_rate": 7.22177742193755e-05,
"loss": 1.1847,
"theoretical_loss": 3.343891101401253,
"tokens_seen": 3063545856
},
{
"epoch": 0.29,
"learning_rate": 7.181745396317054e-05,
"loss": 1.18,
"theoretical_loss": 3.3437825322229298,
"tokens_seen": 3064856576
},
{
"epoch": 0.29,
"learning_rate": 7.141713370696558e-05,
"loss": 1.1817,
"theoretical_loss": 3.3436740224597914,
"tokens_seen": 3066167296
},
{
"epoch": 0.3,
"objective/train/docs_used": 1028068,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.3122754096984863,
"objective/train/theoretical_loss": 3.3435655720539446,
"objective/train/tokens_used": 118762976,
"theoretical_loss": 3.3435655720539446,
"tokens_seen": 3067478016
},
{
"epoch": 0.3,
"learning_rate": 7.101681345076061e-05,
"loss": 1.2002,
"theoretical_loss": 3.3435655720539446,
"tokens_seen": 3067478016
},
{
"epoch": 0.3,
"learning_rate": 7.061649319455565e-05,
"loss": 1.1951,
"theoretical_loss": 3.343457180947577,
"tokens_seen": 3068788736
},
{
"epoch": 0.31,
"learning_rate": 7.021617293835069e-05,
"loss": 1.1923,
"theoretical_loss": 3.343348849082957,
"tokens_seen": 3070099456
},
{
"epoch": 0.31,
"learning_rate": 6.981585268214572e-05,
"loss": 1.1902,
"theoretical_loss": 3.3432405764024353,
"tokens_seen": 3071410176
},
{
"epoch": 0.31,
"learning_rate": 6.941553242594075e-05,
"loss": 1.1889,
"theoretical_loss": 3.3431323628484417,
"tokens_seen": 3072720896
},
{
"epoch": 0.32,
"learning_rate": 6.901521216973579e-05,
"loss": 1.1913,
"theoretical_loss": 3.343024208363487,
"tokens_seen": 3074031616
},
{
"epoch": 0.32,
"learning_rate": 6.861489191353083e-05,
"loss": 1.18,
"theoretical_loss": 3.3429161128901628,
"tokens_seen": 3075342336
},
{
"epoch": 0.32,
"learning_rate": 6.821457165732587e-05,
"loss": 1.1859,
"theoretical_loss": 3.3428080763711403,
"tokens_seen": 3076653056
},
{
"epoch": 0.33,
"learning_rate": 6.78142514011209e-05,
"loss": 1.1819,
"theoretical_loss": 3.3427000987491717,
"tokens_seen": 3077963776
},
{
"epoch": 0.33,
"learning_rate": 6.741393114491593e-05,
"loss": 1.1947,
"theoretical_loss": 3.342592179967088,
"tokens_seen": 3079274496
},
{
"epoch": 0.34,
"learning_rate": 6.701361088871097e-05,
"loss": 1.196,
"theoretical_loss": 3.3424843199678005,
"tokens_seen": 3080585216
},
{
"epoch": 0.34,
"learning_rate": 6.661329063250601e-05,
"loss": 1.1844,
"theoretical_loss": 3.3423765186943006,
"tokens_seen": 3081895936
},
{
"epoch": 0.34,
"learning_rate": 6.621297037630104e-05,
"loss": 1.1972,
"theoretical_loss": 3.3422687760896586,
"tokens_seen": 3083206656
},
{
"epoch": 0.35,
"objective/train/docs_used": 1033433,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.273867130279541,
"objective/train/theoretical_loss": 3.3422149267703904,
"objective/train/tokens_used": 135146976,
"theoretical_loss": 3.3422149267703904,
"tokens_seen": 3083862016
},
{
"epoch": 0.35,
"learning_rate": 6.581265012009608e-05,
"loss": 1.1819,
"theoretical_loss": 3.342161092097024,
"tokens_seen": 3084517376
},
{
"epoch": 0.35,
"learning_rate": 6.541232986389112e-05,
"loss": 1.1736,
"theoretical_loss": 3.342053466659626,
"tokens_seen": 3085828096
},
{
"epoch": 0.36,
"learning_rate": 6.501200960768615e-05,
"loss": 1.1701,
"theoretical_loss": 3.3419458997207734,
"tokens_seen": 3087138816
},
{
"epoch": 0.36,
"learning_rate": 6.461168935148118e-05,
"loss": 1.1776,
"theoretical_loss": 3.3418383912238525,
"tokens_seen": 3088449536
},
{
"epoch": 0.36,
"learning_rate": 6.421136909527622e-05,
"loss": 1.1794,
"theoretical_loss": 3.341730941112329,
"tokens_seen": 3089760256
},
{
"epoch": 0.37,
"learning_rate": 6.381104883907126e-05,
"loss": 1.1969,
"theoretical_loss": 3.341623549329748,
"tokens_seen": 3091070976
},
{
"epoch": 0.37,
"learning_rate": 6.34107285828663e-05,
"loss": 1.1738,
"theoretical_loss": 3.3415162158197327,
"tokens_seen": 3092381696
},
{
"epoch": 0.38,
"learning_rate": 6.301040832666132e-05,
"loss": 1.1815,
"theoretical_loss": 3.3414089405259837,
"tokens_seen": 3093692416
},
{
"epoch": 0.38,
"learning_rate": 6.261008807045636e-05,
"loss": 1.176,
"theoretical_loss": 3.341301723392281,
"tokens_seen": 3095003136
},
{
"epoch": 0.38,
"learning_rate": 6.22097678142514e-05,
"loss": 1.1834,
"theoretical_loss": 3.3411945643624814,
"tokens_seen": 3096313856
},
{
"epoch": 0.39,
"learning_rate": 6.180944755804644e-05,
"loss": 1.1951,
"theoretical_loss": 3.3410874633805214,
"tokens_seen": 3097624576
},
{
"epoch": 0.39,
"learning_rate": 6.140912730184148e-05,
"loss": 1.1936,
"theoretical_loss": 3.3409804203904145,
"tokens_seen": 3098935296
},
{
"epoch": 0.4,
"objective/train/docs_used": 1039074,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.3948986530303955,
"objective/train/theoretical_loss": 3.3408734353362513,
"objective/train/tokens_used": 151530976,
"theoretical_loss": 3.3408734353362513,
"tokens_seen": 3100246016
},
{
"epoch": 0.4,
"learning_rate": 6.1008807045636506e-05,
"loss": 1.1987,
"theoretical_loss": 3.3408734353362513,
"tokens_seen": 3100246016
},
{
"epoch": 0.4,
"learning_rate": 6.060848678943155e-05,
"loss": 1.1874,
"theoretical_loss": 3.3407665081622,
"tokens_seen": 3101556736
},
{
"epoch": 0.4,
"learning_rate": 6.0208166533226584e-05,
"loss": 1.184,
"theoretical_loss": 3.340659638812507,
"tokens_seen": 3102867456
},
{
"epoch": 0.41,
"learning_rate": 5.9807846277021617e-05,
"loss": 1.1919,
"theoretical_loss": 3.340552827231495,
"tokens_seen": 3104178176
},
{
"epoch": 0.41,
"learning_rate": 5.940752602081665e-05,
"loss": 1.1835,
"theoretical_loss": 3.3404460733635646,
"tokens_seen": 3105488896
},
{
"epoch": 0.42,
"learning_rate": 5.9007205764611695e-05,
"loss": 1.18,
"theoretical_loss": 3.340339377153193,
"tokens_seen": 3106799616
},
{
"epoch": 0.42,
"learning_rate": 5.860688550840673e-05,
"loss": 1.1818,
"theoretical_loss": 3.340232738544933,
"tokens_seen": 3108110336
},
{
"epoch": 0.42,
"learning_rate": 5.820656525220176e-05,
"loss": 1.1953,
"theoretical_loss": 3.340126157483417,
"tokens_seen": 3109421056
},
{
"epoch": 0.43,
"learning_rate": 5.7806244995996806e-05,
"loss": 1.1923,
"theoretical_loss": 3.340019633913351,
"tokens_seen": 3110731776
},
{
"epoch": 0.43,
"learning_rate": 5.740592473979184e-05,
"loss": 1.2147,
"theoretical_loss": 3.3399131677795184,
"tokens_seen": 3112042496
},
{
"epoch": 0.44,
"learning_rate": 5.700560448358687e-05,
"loss": 1.2106,
"theoretical_loss": 3.3398067590267804,
"tokens_seen": 3113353216
},
{
"epoch": 0.44,
"learning_rate": 5.66052842273819e-05,
"loss": 1.1884,
"theoretical_loss": 3.339700407600071,
"tokens_seen": 3114663936
},
{
"epoch": 0.44,
"learning_rate": 5.620496397117695e-05,
"loss": 1.186,
"theoretical_loss": 3.339594113444404,
"tokens_seen": 3115974656
},
{
"epoch": 0.45,
"objective/train/docs_used": 1044567,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.3007320165634155,
"objective/train/theoretical_loss": 3.339540987826048,
"objective/train/tokens_used": 167914976,
"theoretical_loss": 3.339540987826048,
"tokens_seen": 3116630016
},
{
"epoch": 0.45,
"learning_rate": 5.580464371497198e-05,
"loss": 1.1914,
"theoretical_loss": 3.3394878765048657,
"tokens_seen": 3117285376
},
{
"epoch": 0.45,
"learning_rate": 5.5404323458767014e-05,
"loss": 1.1877,
"theoretical_loss": 3.3393816967266208,
"tokens_seen": 3118596096
},
{
"epoch": 0.46,
"learning_rate": 5.5004003202562046e-05,
"loss": 1.1988,
"theoretical_loss": 3.3392755740549074,
"tokens_seen": 3119906816
},
{
"epoch": 0.46,
"learning_rate": 5.460368294635709e-05,
"loss": 1.1582,
"theoretical_loss": 3.33916950843504,
"tokens_seen": 3121217536
},
{
"epoch": 0.46,
"learning_rate": 5.4203362690152125e-05,
"loss": 1.174,
"theoretical_loss": 3.3390634998124096,
"tokens_seen": 3122528256
},
{
"epoch": 0.47,
"learning_rate": 5.380304243394716e-05,
"loss": 1.1787,
"theoretical_loss": 3.3389575481324796,
"tokens_seen": 3123838976
},
{
"epoch": 0.47,
"learning_rate": 5.340272217774219e-05,
"loss": 1.2092,
"theoretical_loss": 3.338851653340791,
"tokens_seen": 3125149696
},
{
"epoch": 0.48,
"learning_rate": 5.3002401921537236e-05,
"loss": 1.1857,
"theoretical_loss": 3.3387458153829583,
"tokens_seen": 3126460416
},
{
"epoch": 0.48,
"learning_rate": 5.260208166533227e-05,
"loss": 1.1758,
"theoretical_loss": 3.338640034204672,
"tokens_seen": 3127771136
},
{
"epoch": 0.48,
"learning_rate": 5.22017614091273e-05,
"loss": 1.1887,
"theoretical_loss": 3.3385343097516946,
"tokens_seen": 3129081856
},
{
"epoch": 0.49,
"learning_rate": 5.180144115292234e-05,
"loss": 1.1717,
"theoretical_loss": 3.338428641969867,
"tokens_seen": 3130392576
},
{
"epoch": 0.49,
"learning_rate": 5.140112089671738e-05,
"loss": 1.1977,
"theoretical_loss": 3.3383230308051006,
"tokens_seen": 3131703296
},
{
"debugging/Compilability": 0.0,
"debugging/distinct-1-grams": 0.7006575936203664,
"debugging/entropy-1-grams": 4.797175896655273,
"debugging/length": 597.5,
"debugging/num_segments": 6,
"debugging/score": 0.06411644637926668,
"debugging/score_std": 0.022477740998569338,
"epoch": 0.49,
"objective/train/docs_used": 1050300,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.1647751331329346,
"objective/train/theoretical_loss": 3.338217476203384,
"objective/train/tokens_used": 184298976,
"theoretical_loss": 3.338217476203384,
"tokens_seen": 3133014016
},
{
"epoch": 0.5,
"learning_rate": 5.100080064051241e-05,
"loss": 1.1832,
"theoretical_loss": 3.338217476203384,
"tokens_seen": 3133014016
},
{
"epoch": 0.5,
"learning_rate": 5.060048038430745e-05,
"loss": 1.1799,
"theoretical_loss": 3.3381119781107778,
"tokens_seen": 3134324736
},
{
"epoch": 0.5,
"learning_rate": 5.020016012810248e-05,
"loss": 1.1739,
"theoretical_loss": 3.3380065364734186,
"tokens_seen": 3135635456
},
{
"epoch": 0.51,
"learning_rate": 4.979983987189752e-05,
"loss": 1.164,
"theoretical_loss": 3.3379011512375145,
"tokens_seen": 3136946176
},
{
"epoch": 0.51,
"learning_rate": 4.9399519615692554e-05,
"loss": 1.1768,
"theoretical_loss": 3.3377958223493485,
"tokens_seen": 3138256896
},
{
"epoch": 0.52,
"learning_rate": 4.8999199359487594e-05,
"loss": 1.1813,
"theoretical_loss": 3.337690549755278,
"tokens_seen": 3139567616
},
{
"epoch": 0.52,
"learning_rate": 4.859887910328263e-05,
"loss": 1.1658,
"theoretical_loss": 3.3375853334017327,
"tokens_seen": 3140878336
},
{
"epoch": 0.52,
"learning_rate": 4.8198558847077665e-05,
"loss": 1.2071,
"theoretical_loss": 3.3374801732352157,
"tokens_seen": 3142189056
},
{
"epoch": 0.53,
"learning_rate": 4.7798238590872704e-05,
"loss": 1.1912,
"theoretical_loss": 3.337375069202303,
"tokens_seen": 3143499776
},
{
"epoch": 0.53,
"learning_rate": 4.739791833466774e-05,
"loss": 1.1733,
"theoretical_loss": 3.337270021249646,
"tokens_seen": 3144810496
},
{
"epoch": 0.53,
"learning_rate": 4.6997598078462776e-05,
"loss": 1.1903,
"theoretical_loss": 3.3371650293239643,
"tokens_seen": 3146121216
},
{
"epoch": 0.54,
"learning_rate": 4.659727782225781e-05,
"loss": 1.1848,
"theoretical_loss": 3.337060093372055,
"tokens_seen": 3147431936
},
{
"epoch": 0.54,
"learning_rate": 4.619695756605285e-05,
"loss": 1.1835,
"theoretical_loss": 3.3369552133407856,
"tokens_seen": 3148742656
},
{
"epoch": 0.54,
"objective/train/docs_used": 1056133,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.0949712991714478,
"objective/train/theoretical_loss": 3.336902794278807,
"objective/train/tokens_used": 200682976,
"theoretical_loss": 3.336902794278807,
"tokens_seen": 3149398016
},
{
"epoch": 0.55,
"learning_rate": 4.579663730984788e-05,
"loss": 1.1868,
"theoretical_loss": 3.3368503891770964,
"tokens_seen": 3150053376
},
{
"epoch": 0.55,
"learning_rate": 4.539631705364292e-05,
"loss": 1.1922,
"theoretical_loss": 3.336745620828,
"tokens_seen": 3151364096
},
{
"epoch": 0.55,
"learning_rate": 4.499599679743795e-05,
"loss": 1.2032,
"theoretical_loss": 3.3366409082405823,
"tokens_seen": 3152674816
},
{
"epoch": 0.56,
"learning_rate": 4.459567654123299e-05,
"loss": 1.1804,
"theoretical_loss": 3.3365362513619994,
"tokens_seen": 3153985536
},
{
"epoch": 0.56,
"learning_rate": 4.419535628502802e-05,
"loss": 1.1779,
"theoretical_loss": 3.336431650139481,
"tokens_seen": 3155296256
},
{
"epoch": 0.57,
"learning_rate": 4.379503602882306e-05,
"loss": 1.1845,
"theoretical_loss": 3.336327104520328,
"tokens_seen": 3156606976
},
{
"epoch": 0.57,
"learning_rate": 4.3394715772618095e-05,
"loss": 1.1836,
"theoretical_loss": 3.3362226144519136,
"tokens_seen": 3157917696
},
{
"epoch": 0.57,
"learning_rate": 4.2994395516413134e-05,
"loss": 1.1913,
"theoretical_loss": 3.336118179881682,
"tokens_seen": 3159228416
},
{
"epoch": 0.58,
"learning_rate": 4.2594075260208167e-05,
"loss": 1.1818,
"theoretical_loss": 3.336013800757149,
"tokens_seen": 3160539136
},
{
"epoch": 0.58,
"learning_rate": 4.2193755004003206e-05,
"loss": 1.1748,
"theoretical_loss": 3.3359094770259023,
"tokens_seen": 3161849856
},
{
"epoch": 0.59,
"learning_rate": 4.179343474779824e-05,
"loss": 1.1827,
"theoretical_loss": 3.3358052086356,
"tokens_seen": 3163160576
},
{
"epoch": 0.59,
"learning_rate": 4.139311449159328e-05,
"loss": 1.1664,
"theoretical_loss": 3.3357009955339727,
"tokens_seen": 3164471296
},
{
"epoch": 0.59,
"objective/train/docs_used": 1061123,
"objective/train/instantaneous_batch_size": 16,
"objective/train/instantaneous_microbatch_size": 16384,
"objective/train/original_loss": 1.0366129875183105,
"objective/train/theoretical_loss": 3.33559683766882,
"objective/train/tokens_used": 217066976,
"theoretical_loss": 3.33559683766882,
"tokens_seen": 3165782016
},
{
"epoch": 0.59,
"learning_rate": 4.099279423538831e-05,
"loss": 1.193,
"theoretical_loss": 3.33559683766882,
"tokens_seen": 3165782016
},
{
"epoch": 0.6,
"learning_rate": 4.059247397918335e-05,
"loss": 1.1853,
"theoretical_loss": 3.335492734988014,
"tokens_seen": 3167092736
},
{
"epoch": 0.6,
"learning_rate": 4.019215372297839e-05,
"loss": 1.1756,
"theoretical_loss": 3.3353886874394973,
"tokens_seen": 3168403456
},
{
"epoch": 0.61,
"learning_rate": 3.979183346677342e-05,
"loss": 1.1937,
"theoretical_loss": 3.3352846949712824,
"tokens_seen": 3169714176
},
{
"epoch": 0.61,
"learning_rate": 3.939151321056846e-05,
"loss": 1.183,
"theoretical_loss": 3.3351807575314525,
"tokens_seen": 3171024896
},
{
"epoch": 0.61,
"learning_rate": 3.899119295436349e-05,
"loss": 1.1859,
"theoretical_loss": 3.3350768750681623,
"tokens_seen": 3172335616
},
{
"epoch": 0.62,
"learning_rate": 3.859087269815853e-05,
"loss": 1.2008,
"theoretical_loss": 3.334973047529635,
"tokens_seen": 3173646336
},
{
"epoch": 0.62,
"learning_rate": 3.8190552441953564e-05,
"loss": 1.2086,
"theoretical_loss": 3.3348692748641646,
"tokens_seen": 3174957056
},
{
"epoch": 0.63,
"learning_rate": 3.77902321857486e-05,
"loss": 1.1673,
"theoretical_loss": 3.3347655570201162,
"tokens_seen": 3176267776
},
{
"epoch": 0.63,
"learning_rate": 3.7389911929543636e-05,
"loss": 1.1762,
"theoretical_loss": 3.334661893945923,
"tokens_seen": 3177578496
},
{
"epoch": 0.63,
"learning_rate": 3.6989591673338675e-05,
"loss": 1.165,
"theoretical_loss": 3.3345582855900893,
"tokens_seen": 3178889216
}
],
"max_steps": 2524,
"num_train_epochs": 9223372036854775807,
"total_flos": 1.070252163072e+17,
"trial_name": null,
"trial_params": null
}