naughty_davinci / checkpoint-2300 /trainer_state.json
tomekkorbak's picture
Training in progress, step 2300
0254be1
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9112519809825673,
"global_step": 2300,
"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": 2.7796,
"theoretical_loss": 3.3518567762894107,
"tokens_seen": 2969305088
},
{
"epoch": 0.0,
"learning_rate": 3.846153846153846e-05,
"loss": 2.4642,
"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": 1.4437,
"theoretical_loss": 3.351529257004948,
"tokens_seen": 2973106176
},
{
"epoch": 0.02,
"learning_rate": 9.943955164131305e-05,
"loss": 1.273,
"theoretical_loss": 3.3514164435728655,
"tokens_seen": 2974416896
},
{
"epoch": 0.02,
"learning_rate": 9.90392313851081e-05,
"loss": 1.2074,
"theoretical_loss": 3.3513036937553267,
"tokens_seen": 2975727616
},
{
"epoch": 0.02,
"learning_rate": 9.863891112890312e-05,
"loss": 1.2079,
"theoretical_loss": 3.3511910074884628,
"tokens_seen": 2977038336
},
{
"epoch": 0.02,
"objective/train/docs_used": 997783,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 0.9412798285484314,
"objective/train/theoretical_loss": 3.3511628458440637,
"objective/train/tokens_used": 28650976,
"theoretical_loss": 3.3511628458440637,
"tokens_seen": 2977366016
},
{
"epoch": 0.03,
"learning_rate": 9.823859087269817e-05,
"loss": 1.2032,
"theoretical_loss": 3.3510783847084977,
"tokens_seen": 2978349056
},
{
"epoch": 0.03,
"learning_rate": 9.78382706164932e-05,
"loss": 1.2178,
"theoretical_loss": 3.350965825351748,
"tokens_seen": 2979659776
},
{
"epoch": 0.04,
"learning_rate": 9.743795036028824e-05,
"loss": 1.189,
"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": 1.1797,
"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": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.0221836566925049,
"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": 1.1658,
"theoretical_loss": 3.3504039776936994,
"tokens_seen": 2986213376
},
{
"epoch": 0.06,
"learning_rate": 9.543634907926342e-05,
"loss": 1.1823,
"theoretical_loss": 3.350291797544005,
"tokens_seen": 2987524096
},
{
"epoch": 0.06,
"learning_rate": 9.503602882305846e-05,
"loss": 1.1859,
"theoretical_loss": 3.3501796803742994,
"tokens_seen": 2988834816
},
{
"epoch": 0.06,
"learning_rate": 9.463570856685348e-05,
"loss": 1.1906,
"theoretical_loss": 3.3500676261216285,
"tokens_seen": 2990145536
},
{
"epoch": 0.07,
"learning_rate": 9.423538831064852e-05,
"loss": 1.185,
"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": 0.07,
"objective/train/docs_used": 1002769,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.227379560470581,
"objective/train/theoretical_loss": 3.349759800829844,
"objective/train/tokens_used": 45034976,
"theoretical_loss": 3.349759800829844,
"tokens_seen": 2993750016
},
{
"epoch": 0.08,
"learning_rate": 9.34347477982386e-05,
"loss": 1.1807,
"theoretical_loss": 3.3497318402376397,
"tokens_seen": 2994077696
},
{
"epoch": 0.08,
"learning_rate": 9.303442754203363e-05,
"loss": 1.1862,
"theoretical_loss": 3.349620037025374,
"tokens_seen": 2995388416
},
{
"epoch": 0.08,
"learning_rate": 9.263410728582867e-05,
"loss": 1.178,
"theoretical_loss": 3.349508296416727,
"tokens_seen": 2996699136
},
{
"epoch": 0.09,
"learning_rate": 9.22337870296237e-05,
"loss": 1.1656,
"theoretical_loss": 3.3493966183492847,
"tokens_seen": 2998009856
},
{
"epoch": 0.09,
"learning_rate": 9.183346677341874e-05,
"loss": 1.1769,
"theoretical_loss": 3.349285002760723,
"tokens_seen": 2999320576
},
{
"epoch": 0.1,
"learning_rate": 9.143314651721377e-05,
"loss": 1.1513,
"theoretical_loss": 3.3491734495888066,
"tokens_seen": 3000631296
},
{
"epoch": 0.1,
"objective/train/docs_used": 1005438,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1515272855758667,
"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": 1.1793,
"theoretical_loss": 3.3490619587713906,
"tokens_seen": 3001942016
},
{
"epoch": 0.1,
"learning_rate": 9.063250600480385e-05,
"loss": 1.1534,
"theoretical_loss": 3.348950530246418,
"tokens_seen": 3003252736
},
{
"epoch": 0.11,
"learning_rate": 9.023218574859889e-05,
"loss": 1.1928,
"theoretical_loss": 3.348839163951921,
"tokens_seen": 3004563456
},
{
"epoch": 0.11,
"learning_rate": 8.983186549239391e-05,
"loss": 1.1678,
"theoretical_loss": 3.3487278598260195,
"tokens_seen": 3005874176
},
{
"epoch": 0.11,
"learning_rate": 8.943154523618895e-05,
"loss": 1.1477,
"theoretical_loss": 3.348616617806923,
"tokens_seen": 3007184896
},
{
"epoch": 0.12,
"learning_rate": 8.903122497998399e-05,
"loss": 1.1773,
"theoretical_loss": 3.34850543783293,
"tokens_seen": 3008495616
},
{
"epoch": 0.12,
"learning_rate": 8.863090472377903e-05,
"loss": 1.185,
"theoretical_loss": 3.3483943198424253,
"tokens_seen": 3009806336
},
{
"epoch": 0.12,
"objective/train/docs_used": 1008334,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.3460177183151245,
"objective/train/theoretical_loss": 3.348366550022507,
"objective/train/tokens_used": 61418976,
"theoretical_loss": 3.348366550022507,
"tokens_seen": 3010134016
},
{
"epoch": 0.13,
"learning_rate": 8.823058446757406e-05,
"loss": 1.1798,
"theoretical_loss": 3.348283263773882,
"tokens_seen": 3011117056
},
{
"epoch": 0.13,
"learning_rate": 8.78302642113691e-05,
"loss": 1.1794,
"theoretical_loss": 3.3481722695658633,
"tokens_seen": 3012427776
},
{
"epoch": 0.13,
"learning_rate": 8.742994395516414e-05,
"loss": 1.1759,
"theoretical_loss": 3.3480613371570165,
"tokens_seen": 3013738496
},
{
"epoch": 0.14,
"learning_rate": 8.702962369895917e-05,
"loss": 1.1839,
"theoretical_loss": 3.347950466486079,
"tokens_seen": 3015049216
},
{
"epoch": 0.14,
"learning_rate": 8.66293034427542e-05,
"loss": 1.1996,
"theoretical_loss": 3.3478396574918747,
"tokens_seen": 3016359936
},
{
"epoch": 0.15,
"learning_rate": 8.622898318654924e-05,
"loss": 1.1897,
"theoretical_loss": 3.347728910113315,
"tokens_seen": 3017670656
},
{
"epoch": 0.15,
"objective/train/docs_used": 1011213,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.227089524269104,
"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": 1.1741,
"theoretical_loss": 3.3476182242893975,
"tokens_seen": 3018981376
},
{
"epoch": 0.15,
"learning_rate": 8.542834267413932e-05,
"loss": 1.191,
"theoretical_loss": 3.3475075999592074,
"tokens_seen": 3020292096
},
{
"epoch": 0.16,
"learning_rate": 8.502802241793434e-05,
"loss": 1.1924,
"theoretical_loss": 3.347397037061916,
"tokens_seen": 3021602816
},
{
"epoch": 0.16,
"learning_rate": 8.462770216172938e-05,
"loss": 1.188,
"theoretical_loss": 3.3472865355367825,
"tokens_seen": 3022913536
},
{
"epoch": 0.17,
"learning_rate": 8.422738190552442e-05,
"loss": 1.1754,
"theoretical_loss": 3.3471760953231504,
"tokens_seen": 3024224256
},
{
"epoch": 0.17,
"learning_rate": 8.382706164931946e-05,
"loss": 1.2086,
"theoretical_loss": 3.3470657163604507,
"tokens_seen": 3025534976
},
{
"epoch": 0.17,
"objective/train/docs_used": 1013697,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1135369539260864,
"objective/train/theoretical_loss": 3.3469829722979627,
"objective/train/tokens_used": 77802976,
"theoretical_loss": 3.3469829722979627,
"tokens_seen": 3026518016
},
{
"epoch": 0.17,
"learning_rate": 8.34267413931145e-05,
"loss": 1.1969,
"theoretical_loss": 3.3469553985882,
"tokens_seen": 3026845696
},
{
"epoch": 0.18,
"learning_rate": 8.302642113690953e-05,
"loss": 1.2176,
"theoretical_loss": 3.3468451419460017,
"tokens_seen": 3028156416
},
{
"epoch": 0.18,
"learning_rate": 8.262610088070457e-05,
"loss": 1.188,
"theoretical_loss": 3.3467349463735436,
"tokens_seen": 3029467136
},
{
"epoch": 0.19,
"learning_rate": 8.22257806244996e-05,
"loss": 1.1834,
"theoretical_loss": 3.3466248118105995,
"tokens_seen": 3030777856
},
{
"epoch": 0.19,
"learning_rate": 8.182546036829464e-05,
"loss": 1.1977,
"theoretical_loss": 3.3465147381970293,
"tokens_seen": 3032088576
},
{
"epoch": 0.19,
"learning_rate": 8.142514011208967e-05,
"loss": 1.1889,
"theoretical_loss": 3.346404725472777,
"tokens_seen": 3033399296
},
{
"epoch": 0.2,
"objective/train/docs_used": 1016482,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.4041112661361694,
"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": 1.1983,
"theoretical_loss": 3.3462947735778727,
"tokens_seen": 3034710016
},
{
"epoch": 0.2,
"learning_rate": 8.062449959967975e-05,
"loss": 1.1694,
"theoretical_loss": 3.346184882452431,
"tokens_seen": 3036020736
},
{
"epoch": 0.21,
"learning_rate": 8.022417934347479e-05,
"loss": 1.1871,
"theoretical_loss": 3.3460750520366513,
"tokens_seen": 3037331456
},
{
"epoch": 0.21,
"learning_rate": 7.982385908726981e-05,
"loss": 1.1833,
"theoretical_loss": 3.3459652822708175,
"tokens_seen": 3038642176
},
{
"epoch": 0.21,
"learning_rate": 7.942353883106487e-05,
"loss": 1.1831,
"theoretical_loss": 3.345855573095298,
"tokens_seen": 3039952896
},
{
"epoch": 0.22,
"learning_rate": 7.902321857485989e-05,
"loss": 1.1877,
"theoretical_loss": 3.3457459244505463,
"tokens_seen": 3041263616
},
{
"epoch": 0.22,
"learning_rate": 7.862289831865493e-05,
"loss": 1.2048,
"theoretical_loss": 3.3456363362770984,
"tokens_seen": 3042574336
},
{
"epoch": 0.22,
"objective/train/docs_used": 1019410,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1971296072006226,
"objective/train/theoretical_loss": 3.3456089486754172,
"objective/train/tokens_used": 94186976,
"theoretical_loss": 3.3456089486754172,
"tokens_seen": 3042902016
},
{
"epoch": 0.23,
"learning_rate": 7.822257806244996e-05,
"loss": 1.1685,
"theoretical_loss": 3.345526808515576,
"tokens_seen": 3043885056
},
{
"epoch": 0.23,
"learning_rate": 7.782225780624501e-05,
"loss": 1.1907,
"theoretical_loss": 3.3454173411066828,
"tokens_seen": 3045195776
},
{
"epoch": 0.23,
"learning_rate": 7.742193755004003e-05,
"loss": 1.1949,
"theoretical_loss": 3.345307933991209,
"tokens_seen": 3046506496
},
{
"epoch": 0.24,
"learning_rate": 7.702161729383507e-05,
"loss": 1.1959,
"theoretical_loss": 3.3451985871100254,
"tokens_seen": 3047817216
},
{
"epoch": 0.24,
"learning_rate": 7.662129703763011e-05,
"loss": 1.1728,
"theoretical_loss": 3.345089300404087,
"tokens_seen": 3049127936
},
{
"epoch": 0.25,
"learning_rate": 7.622097678142515e-05,
"loss": 1.1995,
"theoretical_loss": 3.344980073814434,
"tokens_seen": 3050438656
},
{
"debugging/Compilability": 0.0,
"debugging/distinct-1-grams": 0.7613636363636364,
"debugging/entropy-1-grams": 3.9967058025744,
"debugging/length": 568.0,
"debugging/num_segments": 1,
"debugging/score": 0.045774647887323945,
"debugging/score_std": 0.0,
"epoch": 0.25,
"objective/train/docs_used": 1022217,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 0.8801804780960083,
"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": 1.1905,
"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": 1.2223,
"theoretical_loss": 3.3444348405525948,
"tokens_seen": 3056992256
},
{
"epoch": 0.27,
"learning_rate": 7.381905524419536e-05,
"loss": 1.2028,
"theoretical_loss": 3.3443259734271016,
"tokens_seen": 3058302976
},
{
"epoch": 0.27,
"objective/train/docs_used": 1025165,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.2416154146194458,
"objective/train/theoretical_loss": 3.3442443622681504,
"objective/train/tokens_used": 110570976,
"theoretical_loss": 3.3442443622681504,
"tokens_seen": 3059286016
},
{
"epoch": 0.27,
"learning_rate": 7.34187349879904e-05,
"loss": 1.1917,
"theoretical_loss": 3.344217166007481,
"tokens_seen": 3059613696
},
{
"epoch": 0.28,
"learning_rate": 7.301841473178544e-05,
"loss": 1.1963,
"theoretical_loss": 3.344108418235432,
"tokens_seen": 3060924416
},
{
"epoch": 0.28,
"learning_rate": 7.261809447558046e-05,
"loss": 1.1737,
"theoretical_loss": 3.343999730052735,
"tokens_seen": 3062235136
},
{
"epoch": 0.29,
"learning_rate": 7.22177742193755e-05,
"loss": 1.1848,
"theoretical_loss": 3.343891101401253,
"tokens_seen": 3063545856
},
{
"epoch": 0.29,
"learning_rate": 7.181745396317054e-05,
"loss": 1.1799,
"theoretical_loss": 3.3437825322229298,
"tokens_seen": 3064856576
},
{
"epoch": 0.29,
"learning_rate": 7.141713370696558e-05,
"loss": 1.1818,
"theoretical_loss": 3.3436740224597914,
"tokens_seen": 3066167296
},
{
"epoch": 0.3,
"objective/train/docs_used": 1028068,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.3919180631637573,
"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": 1.1999,
"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": 1.1921,
"theoretical_loss": 3.343348849082957,
"tokens_seen": 3070099456
},
{
"epoch": 0.31,
"learning_rate": 6.981585268214572e-05,
"loss": 1.1899,
"theoretical_loss": 3.3432405764024353,
"tokens_seen": 3071410176
},
{
"epoch": 0.31,
"learning_rate": 6.941553242594075e-05,
"loss": 1.1887,
"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": 1.1801,
"theoretical_loss": 3.3429161128901628,
"tokens_seen": 3075342336
},
{
"epoch": 0.32,
"objective/train/docs_used": 1031046,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.355250597000122,
"objective/train/theoretical_loss": 3.342889098235679,
"objective/train/tokens_used": 126954976,
"theoretical_loss": 3.342889098235679,
"tokens_seen": 3075670016
},
{
"epoch": 0.32,
"learning_rate": 6.821457165732587e-05,
"loss": 1.186,
"theoretical_loss": 3.3428080763711403,
"tokens_seen": 3076653056
},
{
"epoch": 0.33,
"learning_rate": 6.78142514011209e-05,
"loss": 1.1817,
"theoretical_loss": 3.3427000987491717,
"tokens_seen": 3077963776
},
{
"epoch": 0.33,
"learning_rate": 6.741393114491593e-05,
"loss": 1.1953,
"theoretical_loss": 3.342592179967088,
"tokens_seen": 3079274496
},
{
"epoch": 0.34,
"learning_rate": 6.701361088871097e-05,
"loss": 1.1954,
"theoretical_loss": 3.3424843199678005,
"tokens_seen": 3080585216
},
{
"epoch": 0.34,
"learning_rate": 6.661329063250601e-05,
"loss": 1.1849,
"theoretical_loss": 3.3423765186943006,
"tokens_seen": 3081895936
},
{
"epoch": 0.34,
"learning_rate": 6.621297037630104e-05,
"loss": 1.1971,
"theoretical_loss": 3.3422687760896586,
"tokens_seen": 3083206656
},
{
"epoch": 0.35,
"objective/train/docs_used": 1033433,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.3183742761611938,
"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": 1.1814,
"theoretical_loss": 3.342161092097024,
"tokens_seen": 3084517376
},
{
"epoch": 0.35,
"learning_rate": 6.541232986389112e-05,
"loss": 1.1732,
"theoretical_loss": 3.342053466659626,
"tokens_seen": 3085828096
},
{
"epoch": 0.36,
"learning_rate": 6.501200960768615e-05,
"loss": 1.1698,
"theoretical_loss": 3.3419458997207734,
"tokens_seen": 3087138816
},
{
"epoch": 0.36,
"learning_rate": 6.461168935148118e-05,
"loss": 1.1775,
"theoretical_loss": 3.3418383912238525,
"tokens_seen": 3088449536
},
{
"epoch": 0.36,
"learning_rate": 6.421136909527622e-05,
"loss": 1.1784,
"theoretical_loss": 3.341730941112329,
"tokens_seen": 3089760256
},
{
"epoch": 0.37,
"learning_rate": 6.381104883907126e-05,
"loss": 1.1965,
"theoretical_loss": 3.341623549329748,
"tokens_seen": 3091070976
},
{
"epoch": 0.37,
"objective/train/docs_used": 1036380,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.2508410215377808,
"objective/train/theoretical_loss": 3.3415430437372637,
"objective/train/tokens_used": 143338976,
"theoretical_loss": 3.3415430437372637,
"tokens_seen": 3092054016
},
{
"epoch": 0.37,
"learning_rate": 6.34107285828663e-05,
"loss": 1.1741,
"theoretical_loss": 3.3415162158197327,
"tokens_seen": 3092381696
},
{
"epoch": 0.38,
"learning_rate": 6.301040832666132e-05,
"loss": 1.1813,
"theoretical_loss": 3.3414089405259837,
"tokens_seen": 3093692416
},
{
"epoch": 0.38,
"learning_rate": 6.261008807045636e-05,
"loss": 1.1763,
"theoretical_loss": 3.341301723392281,
"tokens_seen": 3095003136
},
{
"epoch": 0.38,
"learning_rate": 6.22097678142514e-05,
"loss": 1.1828,
"theoretical_loss": 3.3411945643624814,
"tokens_seen": 3096313856
},
{
"epoch": 0.39,
"learning_rate": 6.180944755804644e-05,
"loss": 1.1958,
"theoretical_loss": 3.3410874633805214,
"tokens_seen": 3097624576
},
{
"epoch": 0.39,
"learning_rate": 6.140912730184148e-05,
"loss": 1.1942,
"theoretical_loss": 3.3409804203904145,
"tokens_seen": 3098935296
},
{
"epoch": 0.4,
"objective/train/docs_used": 1039074,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.3202261924743652,
"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": 1.1996,
"theoretical_loss": 3.3408734353362513,
"tokens_seen": 3100246016
},
{
"epoch": 0.4,
"learning_rate": 6.060848678943155e-05,
"loss": 1.1876,
"theoretical_loss": 3.3407665081622,
"tokens_seen": 3101556736
},
{
"epoch": 0.4,
"learning_rate": 6.0208166533226584e-05,
"loss": 1.1839,
"theoretical_loss": 3.340659638812507,
"tokens_seen": 3102867456
},
{
"epoch": 0.41,
"learning_rate": 5.9807846277021617e-05,
"loss": 1.192,
"theoretical_loss": 3.340552827231495,
"tokens_seen": 3104178176
},
{
"epoch": 0.41,
"learning_rate": 5.940752602081665e-05,
"loss": 1.1838,
"theoretical_loss": 3.3404460733635646,
"tokens_seen": 3105488896
},
{
"epoch": 0.42,
"learning_rate": 5.9007205764611695e-05,
"loss": 1.1796,
"theoretical_loss": 3.340339377153193,
"tokens_seen": 3106799616
},
{
"epoch": 0.42,
"learning_rate": 5.860688550840673e-05,
"loss": 1.1824,
"theoretical_loss": 3.340232738544933,
"tokens_seen": 3108110336
},
{
"epoch": 0.42,
"objective/train/docs_used": 1042051,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.5605241060256958,
"objective/train/theoretical_loss": 3.340206087886708,
"objective/train/tokens_used": 159722976,
"theoretical_loss": 3.340206087886708,
"tokens_seen": 3108438016
},
{
"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": 1.1918,
"theoretical_loss": 3.340019633913351,
"tokens_seen": 3110731776
},
{
"epoch": 0.43,
"learning_rate": 5.740592473979184e-05,
"loss": 1.2149,
"theoretical_loss": 3.3399131677795184,
"tokens_seen": 3112042496
},
{
"epoch": 0.44,
"learning_rate": 5.700560448358687e-05,
"loss": 1.2105,
"theoretical_loss": 3.3398067590267804,
"tokens_seen": 3113353216
},
{
"epoch": 0.44,
"learning_rate": 5.66052842273819e-05,
"loss": 1.1876,
"theoretical_loss": 3.339700407600071,
"tokens_seen": 3114663936
},
{
"epoch": 0.44,
"learning_rate": 5.620496397117695e-05,
"loss": 1.1859,
"theoretical_loss": 3.339594113444404,
"tokens_seen": 3115974656
},
{
"epoch": 0.45,
"objective/train/docs_used": 1044567,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1618666648864746,
"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": 1.1922,
"theoretical_loss": 3.3394878765048657,
"tokens_seen": 3117285376
},
{
"epoch": 0.45,
"learning_rate": 5.5404323458767014e-05,
"loss": 1.1874,
"theoretical_loss": 3.3393816967266208,
"tokens_seen": 3118596096
},
{
"epoch": 0.46,
"learning_rate": 5.5004003202562046e-05,
"loss": 1.1985,
"theoretical_loss": 3.3392755740549074,
"tokens_seen": 3119906816
},
{
"epoch": 0.46,
"learning_rate": 5.460368294635709e-05,
"loss": 1.1583,
"theoretical_loss": 3.33916950843504,
"tokens_seen": 3121217536
},
{
"epoch": 0.46,
"learning_rate": 5.4203362690152125e-05,
"loss": 1.1743,
"theoretical_loss": 3.3390634998124096,
"tokens_seen": 3122528256
},
{
"epoch": 0.47,
"learning_rate": 5.380304243394716e-05,
"loss": 1.1793,
"theoretical_loss": 3.3389575481324796,
"tokens_seen": 3123838976
},
{
"epoch": 0.47,
"objective/train/docs_used": 1047326,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.2724859714508057,
"objective/train/theoretical_loss": 3.338878121708416,
"objective/train/tokens_used": 176106976,
"theoretical_loss": 3.338878121708416,
"tokens_seen": 3124822016
},
{
"epoch": 0.47,
"learning_rate": 5.340272217774219e-05,
"loss": 1.209,
"theoretical_loss": 3.338851653340791,
"tokens_seen": 3125149696
},
{
"epoch": 0.48,
"learning_rate": 5.3002401921537236e-05,
"loss": 1.1861,
"theoretical_loss": 3.3387458153829583,
"tokens_seen": 3126460416
},
{
"epoch": 0.48,
"learning_rate": 5.260208166533227e-05,
"loss": 1.176,
"theoretical_loss": 3.338640034204672,
"tokens_seen": 3127771136
},
{
"epoch": 0.48,
"learning_rate": 5.22017614091273e-05,
"loss": 1.1889,
"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": 0.6651816214290593,
"debugging/entropy-1-grams": 4.2985685702692615,
"debugging/length": 583.75,
"debugging/num_segments": 4,
"debugging/score": 0.06321276390574523,
"debugging/score_std": 0.02740587124120667,
"epoch": 0.49,
"objective/train/docs_used": 1050300,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.2621127367019653,
"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": 1.1834,
"theoretical_loss": 3.338217476203384,
"tokens_seen": 3133014016
},
{
"epoch": 0.5,
"learning_rate": 5.060048038430745e-05,
"loss": 1.1796,
"theoretical_loss": 3.3381119781107778,
"tokens_seen": 3134324736
},
{
"epoch": 0.5,
"learning_rate": 5.020016012810248e-05,
"loss": 1.1743,
"theoretical_loss": 3.3380065364734186,
"tokens_seen": 3135635456
},
{
"epoch": 0.51,
"learning_rate": 4.979983987189752e-05,
"loss": 1.1641,
"theoretical_loss": 3.3379011512375145,
"tokens_seen": 3136946176
},
{
"epoch": 0.51,
"learning_rate": 4.9399519615692554e-05,
"loss": 1.1767,
"theoretical_loss": 3.3377958223493485,
"tokens_seen": 3138256896
},
{
"epoch": 0.52,
"learning_rate": 4.8999199359487594e-05,
"loss": 1.1809,
"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,
"objective/train/docs_used": 1053223,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.286133885383606,
"objective/train/theoretical_loss": 3.337559038094658,
"objective/train/tokens_used": 192490976,
"theoretical_loss": 3.337559038094658,
"tokens_seen": 3141206016
},
{
"epoch": 0.52,
"learning_rate": 4.8198558847077665e-05,
"loss": 1.2072,
"theoretical_loss": 3.3374801732352157,
"tokens_seen": 3142189056
},
{
"epoch": 0.53,
"learning_rate": 4.7798238590872704e-05,
"loss": 1.1917,
"theoretical_loss": 3.337375069202303,
"tokens_seen": 3143499776
},
{
"epoch": 0.53,
"learning_rate": 4.739791833466774e-05,
"loss": 1.1742,
"theoretical_loss": 3.337270021249646,
"tokens_seen": 3144810496
},
{
"epoch": 0.53,
"learning_rate": 4.6997598078462776e-05,
"loss": 1.1901,
"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": 1.184,
"theoretical_loss": 3.3369552133407856,
"tokens_seen": 3148742656
},
{
"epoch": 0.54,
"objective/train/docs_used": 1056133,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.0991315841674805,
"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": 1.1872,
"theoretical_loss": 3.3368503891770964,
"tokens_seen": 3150053376
},
{
"epoch": 0.55,
"learning_rate": 4.539631705364292e-05,
"loss": 1.1918,
"theoretical_loss": 3.336745620828,
"tokens_seen": 3151364096
},
{
"epoch": 0.55,
"learning_rate": 4.499599679743795e-05,
"loss": 1.2037,
"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": 1.178,
"theoretical_loss": 3.336431650139481,
"tokens_seen": 3155296256
},
{
"epoch": 0.57,
"learning_rate": 4.379503602882306e-05,
"loss": 1.1858,
"theoretical_loss": 3.336327104520328,
"tokens_seen": 3156606976
},
{
"epoch": 0.57,
"objective/train/docs_used": 1058478,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1500177383422852,
"objective/train/theoretical_loss": 3.336248731764011,
"objective/train/tokens_used": 208874976,
"theoretical_loss": 3.336248731764011,
"tokens_seen": 3157590016
},
{
"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": 1.1917,
"theoretical_loss": 3.336118179881682,
"tokens_seen": 3159228416
},
{
"epoch": 0.58,
"learning_rate": 4.2594075260208167e-05,
"loss": 1.1829,
"theoretical_loss": 3.336013800757149,
"tokens_seen": 3160539136
},
{
"epoch": 0.58,
"learning_rate": 4.2193755004003206e-05,
"loss": 1.1744,
"theoretical_loss": 3.3359094770259023,
"tokens_seen": 3161849856
},
{
"epoch": 0.59,
"learning_rate": 4.179343474779824e-05,
"loss": 1.1823,
"theoretical_loss": 3.3358052086356,
"tokens_seen": 3163160576
},
{
"epoch": 0.59,
"learning_rate": 4.139311449159328e-05,
"loss": 1.1667,
"theoretical_loss": 3.3357009955339727,
"tokens_seen": 3164471296
},
{
"epoch": 0.59,
"objective/train/docs_used": 1061123,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 0.7507131099700928,
"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": 1.1938,
"theoretical_loss": 3.33559683766882,
"tokens_seen": 3165782016
},
{
"epoch": 0.6,
"learning_rate": 4.059247397918335e-05,
"loss": 1.1858,
"theoretical_loss": 3.335492734988014,
"tokens_seen": 3167092736
},
{
"epoch": 0.6,
"learning_rate": 4.019215372297839e-05,
"loss": 1.1758,
"theoretical_loss": 3.3353886874394973,
"tokens_seen": 3168403456
},
{
"epoch": 0.61,
"learning_rate": 3.979183346677342e-05,
"loss": 1.1935,
"theoretical_loss": 3.3352846949712824,
"tokens_seen": 3169714176
},
{
"epoch": 0.61,
"learning_rate": 3.939151321056846e-05,
"loss": 1.1835,
"theoretical_loss": 3.3351807575314525,
"tokens_seen": 3171024896
},
{
"epoch": 0.61,
"learning_rate": 3.899119295436349e-05,
"loss": 1.1861,
"theoretical_loss": 3.3350768750681623,
"tokens_seen": 3172335616
},
{
"epoch": 0.62,
"learning_rate": 3.859087269815853e-05,
"loss": 1.2016,
"theoretical_loss": 3.334973047529635,
"tokens_seen": 3173646336
},
{
"epoch": 0.62,
"objective/train/docs_used": 1063909,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.3667196035385132,
"objective/train/theoretical_loss": 3.3349470992209365,
"objective/train/tokens_used": 225258976,
"theoretical_loss": 3.3349470992209365,
"tokens_seen": 3173974016
},
{
"epoch": 0.62,
"learning_rate": 3.8190552441953564e-05,
"loss": 1.208,
"theoretical_loss": 3.3348692748641646,
"tokens_seen": 3174957056
},
{
"epoch": 0.63,
"learning_rate": 3.77902321857486e-05,
"loss": 1.1676,
"theoretical_loss": 3.3347655570201162,
"tokens_seen": 3176267776
},
{
"epoch": 0.63,
"learning_rate": 3.7389911929543636e-05,
"loss": 1.176,
"theoretical_loss": 3.334661893945923,
"tokens_seen": 3177578496
},
{
"epoch": 0.63,
"learning_rate": 3.6989591673338675e-05,
"loss": 1.1655,
"theoretical_loss": 3.3345582855900893,
"tokens_seen": 3178889216
},
{
"epoch": 0.64,
"learning_rate": 3.658927141713371e-05,
"loss": 1.2031,
"theoretical_loss": 3.334454731901188,
"tokens_seen": 3180199936
},
{
"epoch": 0.64,
"learning_rate": 3.6188951160928746e-05,
"loss": 1.1675,
"theoretical_loss": 3.3343512328278626,
"tokens_seen": 3181510656
},
{
"epoch": 0.64,
"objective/train/docs_used": 1066709,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.2202755212783813,
"objective/train/theoretical_loss": 3.3342995037560104,
"objective/train/tokens_used": 233450976,
"theoretical_loss": 3.3342995037560104,
"tokens_seen": 3182166016
},
{
"epoch": 0.65,
"learning_rate": 3.578863090472378e-05,
"loss": 1.1818,
"theoretical_loss": 3.3342477883188244,
"tokens_seen": 3182821376
},
{
"epoch": 0.65,
"learning_rate": 3.538831064851882e-05,
"loss": 1.1759,
"theoretical_loss": 3.334144398322856,
"tokens_seen": 3184132096
},
{
"epoch": 0.65,
"learning_rate": 3.498799039231385e-05,
"loss": 1.199,
"theoretical_loss": 3.3340410627888075,
"tokens_seen": 3185442816
},
{
"epoch": 0.66,
"learning_rate": 3.458767013610889e-05,
"loss": 1.1651,
"theoretical_loss": 3.3339377816655986,
"tokens_seen": 3186753536
},
{
"epoch": 0.66,
"learning_rate": 3.418734987990392e-05,
"loss": 1.1638,
"theoretical_loss": 3.333834554902218,
"tokens_seen": 3188064256
},
{
"epoch": 0.67,
"learning_rate": 3.378702962369896e-05,
"loss": 1.1793,
"theoretical_loss": 3.333731382447723,
"tokens_seen": 3189374976
},
{
"epoch": 0.67,
"objective/train/docs_used": 1069727,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1584644317626953,
"objective/train/theoretical_loss": 3.3336540387164515,
"objective/train/tokens_used": 241642976,
"theoretical_loss": 3.3336540387164515,
"tokens_seen": 3190358016
},
{
"epoch": 0.67,
"learning_rate": 3.3386709367493994e-05,
"loss": 1.1984,
"theoretical_loss": 3.3336282642512396,
"tokens_seen": 3190685696
},
{
"epoch": 0.67,
"learning_rate": 3.298638911128903e-05,
"loss": 1.2295,
"theoretical_loss": 3.3335252002619624,
"tokens_seen": 3191996416
},
{
"epoch": 0.68,
"learning_rate": 3.2586068855084065e-05,
"loss": 1.1866,
"theoretical_loss": 3.3334221904291543,
"tokens_seen": 3193307136
},
{
"epoch": 0.68,
"learning_rate": 3.2185748598879104e-05,
"loss": 1.2311,
"theoretical_loss": 3.3333192347021465,
"tokens_seen": 3194617856
},
{
"epoch": 0.69,
"learning_rate": 3.178542834267414e-05,
"loss": 1.1745,
"theoretical_loss": 3.3332163330303386,
"tokens_seen": 3195928576
},
{
"epoch": 0.69,
"learning_rate": 3.1385108086469176e-05,
"loss": 1.1539,
"theoretical_loss": 3.3331134853631976,
"tokens_seen": 3197239296
},
{
"epoch": 0.69,
"objective/train/docs_used": 1072138,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.353588342666626,
"objective/train/theoretical_loss": 3.3330106916502595,
"objective/train/tokens_used": 249834976,
"theoretical_loss": 3.3330106916502595,
"tokens_seen": 3198550016
},
{
"epoch": 0.69,
"learning_rate": 3.098478783026421e-05,
"loss": 1.1742,
"theoretical_loss": 3.3330106916502595,
"tokens_seen": 3198550016
},
{
"epoch": 0.7,
"learning_rate": 3.058446757405925e-05,
"loss": 1.1751,
"theoretical_loss": 3.332907951841127,
"tokens_seen": 3199860736
},
{
"epoch": 0.7,
"learning_rate": 3.0184147317854284e-05,
"loss": 1.1712,
"theoretical_loss": 3.332805265885471,
"tokens_seen": 3201171456
},
{
"epoch": 0.71,
"learning_rate": 2.9783827061649323e-05,
"loss": 1.1948,
"theoretical_loss": 3.33270263373303,
"tokens_seen": 3202482176
},
{
"epoch": 0.71,
"learning_rate": 2.9383506805444355e-05,
"loss": 1.2016,
"theoretical_loss": 3.33260005533361,
"tokens_seen": 3203792896
},
{
"epoch": 0.71,
"learning_rate": 2.8983186549239394e-05,
"loss": 1.2092,
"theoretical_loss": 3.3324975306370845,
"tokens_seen": 3205103616
},
{
"epoch": 0.72,
"learning_rate": 2.8582866293034427e-05,
"loss": 1.2189,
"theoretical_loss": 3.3323950595933933,
"tokens_seen": 3206414336
},
{
"epoch": 0.72,
"objective/train/docs_used": 1074937,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.3788819313049316,
"objective/train/theoretical_loss": 3.332369450209866,
"objective/train/tokens_used": 258026976,
"theoretical_loss": 3.332369450209866,
"tokens_seen": 3206742016
},
{
"epoch": 0.72,
"learning_rate": 2.8182546036829466e-05,
"loss": 1.1972,
"theoretical_loss": 3.3322926421525447,
"tokens_seen": 3207725056
},
{
"epoch": 0.73,
"learning_rate": 2.77822257806245e-05,
"loss": 1.2114,
"theoretical_loss": 3.332190278264613,
"tokens_seen": 3209035776
},
{
"epoch": 0.73,
"learning_rate": 2.7381905524419538e-05,
"loss": 1.1867,
"theoretical_loss": 3.3320879678797395,
"tokens_seen": 3210346496
},
{
"epoch": 0.73,
"learning_rate": 2.698158526821457e-05,
"loss": 1.1942,
"theoretical_loss": 3.3319857109481323,
"tokens_seen": 3211657216
},
{
"epoch": 0.74,
"learning_rate": 2.658126501200961e-05,
"loss": 1.2116,
"theoretical_loss": 3.331883507420067,
"tokens_seen": 3212967936
},
{
"epoch": 0.74,
"learning_rate": 2.618094475580464e-05,
"loss": 1.2044,
"theoretical_loss": 3.3317813572458834,
"tokens_seen": 3214278656
},
{
"debugging/Compilability": 0.0,
"debugging/distinct-1-grams": 0.7649685585732098,
"debugging/entropy-1-grams": 4.842071493424633,
"debugging/length": 542.8,
"debugging/num_segments": 5,
"debugging/score": 0.05065775917820918,
"debugging/score_std": 0.023966644867425342,
"epoch": 0.74,
"objective/train/docs_used": 1077839,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.1823174953460693,
"objective/train/theoretical_loss": 3.331730302150998,
"objective/train/tokens_used": 266218976,
"theoretical_loss": 3.331730302150998,
"tokens_seen": 3214934016
},
{
"epoch": 0.74,
"learning_rate": 2.578062449959968e-05,
"loss": 1.1801,
"theoretical_loss": 3.331679260375991,
"tokens_seen": 3215589376
},
{
"epoch": 0.75,
"learning_rate": 2.5380304243394713e-05,
"loss": 1.1625,
"theoretical_loss": 3.331577216760862,
"tokens_seen": 3216900096
},
{
"epoch": 0.75,
"learning_rate": 2.4979983987189752e-05,
"loss": 1.2035,
"theoretical_loss": 3.331475226351038,
"tokens_seen": 3218210816
},
{
"epoch": 0.76,
"learning_rate": 2.4579663730984788e-05,
"loss": 1.1689,
"theoretical_loss": 3.331373289097125,
"tokens_seen": 3219521536
},
{
"epoch": 0.76,
"learning_rate": 2.4179343474779824e-05,
"loss": 1.1771,
"theoretical_loss": 3.3312714049497947,
"tokens_seen": 3220832256
},
{
"epoch": 0.76,
"learning_rate": 2.377902321857486e-05,
"loss": 1.2113,
"theoretical_loss": 3.3311695738597855,
"tokens_seen": 3222142976
},
{
"epoch": 0.77,
"objective/train/docs_used": 1080652,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.0484637022018433,
"objective/train/theoretical_loss": 3.3310932353315486,
"objective/train/tokens_used": 274410976,
"theoretical_loss": 3.3310932353315486,
"tokens_seen": 3223126016
},
{
"epoch": 0.77,
"learning_rate": 2.3378702962369896e-05,
"loss": 1.1704,
"theoretical_loss": 3.3310677957779014,
"tokens_seen": 3223453696
},
{
"epoch": 0.77,
"learning_rate": 2.297838270616493e-05,
"loss": 1.2005,
"theoretical_loss": 3.330966070655011,
"tokens_seen": 3224764416
},
{
"epoch": 0.78,
"learning_rate": 2.257806244995997e-05,
"loss": 1.1869,
"theoretical_loss": 3.330864398442049,
"tokens_seen": 3226075136
},
{
"epoch": 0.78,
"learning_rate": 2.2177742193755007e-05,
"loss": 1.1755,
"theoretical_loss": 3.3307627790900165,
"tokens_seen": 3227385856
},
{
"epoch": 0.78,
"learning_rate": 2.1777421937550042e-05,
"loss": 1.1833,
"theoretical_loss": 3.330661212549978,
"tokens_seen": 3228696576
},
{
"epoch": 0.79,
"learning_rate": 2.1377101681345078e-05,
"loss": 1.21,
"theoretical_loss": 3.330559698773065,
"tokens_seen": 3230007296
},
{
"epoch": 0.79,
"objective/train/docs_used": 1083207,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.077799677848816,
"objective/train/theoretical_loss": 3.330458237710472,
"objective/train/tokens_used": 282602976,
"theoretical_loss": 3.330458237710472,
"tokens_seen": 3231318016
},
{
"epoch": 0.79,
"learning_rate": 2.0976781425140114e-05,
"loss": 1.2286,
"theoretical_loss": 3.330458237710472,
"tokens_seen": 3231318016
},
{
"epoch": 0.8,
"learning_rate": 2.057646116893515e-05,
"loss": 1.2068,
"theoretical_loss": 3.3303568293134607,
"tokens_seen": 3232628736
},
{
"epoch": 0.8,
"learning_rate": 2.0176140912730186e-05,
"loss": 1.2013,
"theoretical_loss": 3.3302554735333554,
"tokens_seen": 3233939456
},
{
"epoch": 0.8,
"learning_rate": 1.977582065652522e-05,
"loss": 1.1972,
"theoretical_loss": 3.3301541703215465,
"tokens_seen": 3235250176
},
{
"epoch": 0.81,
"learning_rate": 1.9375500400320257e-05,
"loss": 1.1842,
"theoretical_loss": 3.330052919629489,
"tokens_seen": 3236560896
},
{
"epoch": 0.81,
"learning_rate": 1.8975180144115293e-05,
"loss": 1.214,
"theoretical_loss": 3.329951721408701,
"tokens_seen": 3237871616
},
{
"epoch": 0.82,
"learning_rate": 1.857485988791033e-05,
"loss": 1.1823,
"theoretical_loss": 3.3298505756107666,
"tokens_seen": 3239182336
},
{
"epoch": 0.82,
"objective/train/docs_used": 1085940,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 0.996847927570343,
"objective/train/theoretical_loss": 3.3298252973466864,
"objective/train/tokens_used": 290794976,
"theoretical_loss": 3.3298252973466864,
"tokens_seen": 3239510016
},
{
"epoch": 0.82,
"learning_rate": 1.8174539631705365e-05,
"loss": 1.1685,
"theoretical_loss": 3.3297494821873337,
"tokens_seen": 3240493056
},
{
"epoch": 0.82,
"learning_rate": 1.77742193755004e-05,
"loss": 1.1904,
"theoretical_loss": 3.3296484410901135,
"tokens_seen": 3241803776
},
{
"epoch": 0.83,
"learning_rate": 1.7373899119295436e-05,
"loss": 1.1923,
"theoretical_loss": 3.329547452270882,
"tokens_seen": 3243114496
},
{
"epoch": 0.83,
"learning_rate": 1.6973578863090472e-05,
"loss": 1.1947,
"theoretical_loss": 3.3294465156814788,
"tokens_seen": 3244425216
},
{
"epoch": 0.84,
"learning_rate": 1.6573258606885508e-05,
"loss": 1.198,
"theoretical_loss": 3.3293456312738074,
"tokens_seen": 3245735936
},
{
"epoch": 0.84,
"learning_rate": 1.6172938350680544e-05,
"loss": 1.1954,
"theoretical_loss": 3.3292447989998353,
"tokens_seen": 3247046656
},
{
"epoch": 0.84,
"objective/train/docs_used": 1088712,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.440091848373413,
"objective/train/theoretical_loss": 3.3291944023979934,
"objective/train/tokens_used": 298986976,
"theoretical_loss": 3.3291944023979934,
"tokens_seen": 3247702016
},
{
"epoch": 0.84,
"learning_rate": 1.577261809447558e-05,
"loss": 1.1908,
"theoretical_loss": 3.329144018811593,
"tokens_seen": 3248357376
},
{
"epoch": 0.85,
"learning_rate": 1.5372297838270615e-05,
"loss": 1.2103,
"theoretical_loss": 3.3290432906611747,
"tokens_seen": 3249668096
},
{
"epoch": 0.85,
"learning_rate": 1.4971977582065653e-05,
"loss": 1.1834,
"theoretical_loss": 3.328942614500739,
"tokens_seen": 3250978816
},
{
"epoch": 0.86,
"learning_rate": 1.4571657325860689e-05,
"loss": 1.187,
"theoretical_loss": 3.328841990282505,
"tokens_seen": 3252289536
},
{
"epoch": 0.86,
"learning_rate": 1.4171337069655724e-05,
"loss": 1.1827,
"theoretical_loss": 3.3287414179587578,
"tokens_seen": 3253600256
},
{
"epoch": 0.86,
"learning_rate": 1.377101681345076e-05,
"loss": 1.2075,
"theoretical_loss": 3.3286408974818444,
"tokens_seen": 3254910976
},
{
"epoch": 0.87,
"objective/train/docs_used": 1091411,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.2068538665771484,
"objective/train/theoretical_loss": 3.3285655411200135,
"objective/train/tokens_used": 307178976,
"theoretical_loss": 3.3285655411200135,
"tokens_seen": 3255894016
},
{
"epoch": 0.87,
"learning_rate": 1.3370696557245796e-05,
"loss": 1.1996,
"theoretical_loss": 3.328540428804175,
"tokens_seen": 3256221696
},
{
"epoch": 0.87,
"learning_rate": 1.2970376301040832e-05,
"loss": 1.2049,
"theoretical_loss": 3.3284400118782216,
"tokens_seen": 3257532416
},
{
"epoch": 0.88,
"learning_rate": 1.2570056044835871e-05,
"loss": 1.1909,
"theoretical_loss": 3.3283396466565205,
"tokens_seen": 3258843136
},
{
"epoch": 0.88,
"learning_rate": 1.2169735788630905e-05,
"loss": 1.2063,
"theoretical_loss": 3.3282393330916693,
"tokens_seen": 3260153856
},
{
"epoch": 0.88,
"learning_rate": 1.1769415532425941e-05,
"loss": 1.2061,
"theoretical_loss": 3.3281390711363286,
"tokens_seen": 3261464576
},
{
"epoch": 0.89,
"learning_rate": 1.1369095276220977e-05,
"loss": 1.2009,
"theoretical_loss": 3.328038860743222,
"tokens_seen": 3262775296
},
{
"epoch": 0.89,
"objective/train/docs_used": 1094282,
"objective/train/instantaneous_batch_size": 8,
"objective/train/instantaneous_microbatch_size": 8192,
"objective/train/original_loss": 1.0986521244049072,
"objective/train/theoretical_loss": 3.3279387018651336,
"objective/train/tokens_used": 315370976,
"theoretical_loss": 3.3279387018651336,
"tokens_seen": 3264086016
},
{
"epoch": 0.89,
"learning_rate": 1.0968775020016013e-05,
"loss": 1.1938,
"theoretical_loss": 3.3279387018651336,
"tokens_seen": 3264086016
},
{
"epoch": 0.9,
"learning_rate": 1.0568454763811048e-05,
"loss": 1.1947,
"theoretical_loss": 3.3278385944549114,
"tokens_seen": 3265396736
},
{
"epoch": 0.9,
"learning_rate": 1.0168134507606084e-05,
"loss": 1.1832,
"theoretical_loss": 3.327738538465465,
"tokens_seen": 3266707456
},
{
"epoch": 0.9,
"learning_rate": 9.76781425140112e-06,
"loss": 1.212,
"theoretical_loss": 3.327638533849766,
"tokens_seen": 3268018176
},
{
"epoch": 0.91,
"learning_rate": 9.367493995196158e-06,
"loss": 1.1694,
"theoretical_loss": 3.3275385805608475,
"tokens_seen": 3269328896
},
{
"epoch": 0.91,
"learning_rate": 8.967173738991193e-06,
"loss": 1.197,
"theoretical_loss": 3.327438678551805,
"tokens_seen": 3270639616
}
],
"max_steps": 2524,
"num_train_epochs": 9223372036854775807,
"total_flos": 1.538487484416e+17,
"trial_name": null,
"trial_params": null
}