Chiung-Yi's picture
Upload folder using huggingface_hub
a8aacfb verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 315,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009523809523809525,
"grad_norm": 2.447430372238159,
"learning_rate": 0.0,
"loss": 1.8171,
"memory/device_mem_reserved(gib)": 36.42,
"memory/max_mem_active(gib)": 35.17,
"memory/max_mem_allocated(gib)": 35.17,
"step": 1
},
{
"epoch": 0.01904761904761905,
"grad_norm": 2.280487537384033,
"learning_rate": 2.2222222222222223e-05,
"loss": 1.7544,
"memory/device_mem_reserved(gib)": 36.45,
"memory/max_mem_active(gib)": 35.23,
"memory/max_mem_allocated(gib)": 35.23,
"step": 2
},
{
"epoch": 0.02857142857142857,
"grad_norm": 2.248966693878174,
"learning_rate": 4.4444444444444447e-05,
"loss": 1.7546,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 3
},
{
"epoch": 0.0380952380952381,
"grad_norm": 2.2559540271759033,
"learning_rate": 6.666666666666667e-05,
"loss": 1.8011,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 4
},
{
"epoch": 0.047619047619047616,
"grad_norm": 2.149872064590454,
"learning_rate": 8.888888888888889e-05,
"loss": 1.6465,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 5
},
{
"epoch": 0.05714285714285714,
"grad_norm": 1.6111228466033936,
"learning_rate": 0.00011111111111111112,
"loss": 1.5935,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 6
},
{
"epoch": 0.06666666666666667,
"grad_norm": 1.363952875137329,
"learning_rate": 0.00013333333333333334,
"loss": 1.4371,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 7
},
{
"epoch": 0.0761904761904762,
"grad_norm": 1.1439425945281982,
"learning_rate": 0.00015555555555555556,
"loss": 1.2409,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 8
},
{
"epoch": 0.08571428571428572,
"grad_norm": 0.877252995967865,
"learning_rate": 0.00017777777777777779,
"loss": 1.1083,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 9
},
{
"epoch": 0.09523809523809523,
"grad_norm": 0.8071479797363281,
"learning_rate": 0.0002,
"loss": 0.9937,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 10
},
{
"epoch": 0.10476190476190476,
"grad_norm": 0.7772606015205383,
"learning_rate": 0.00019999472984871732,
"loss": 0.9235,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 11
},
{
"epoch": 0.11428571428571428,
"grad_norm": 0.811764657497406,
"learning_rate": 0.00019997891995035912,
"loss": 0.7635,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 12
},
{
"epoch": 0.12380952380952381,
"grad_norm": 0.8423659801483154,
"learning_rate": 0.0001999525719713366,
"loss": 0.6836,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 13
},
{
"epoch": 0.13333333333333333,
"grad_norm": 0.7858280539512634,
"learning_rate": 0.0001999156886888064,
"loss": 0.6159,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 14
},
{
"epoch": 0.14285714285714285,
"grad_norm": 0.739560604095459,
"learning_rate": 0.00019986827399037812,
"loss": 0.5082,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 15
},
{
"epoch": 0.1523809523809524,
"grad_norm": 0.6787500381469727,
"learning_rate": 0.00019981033287370443,
"loss": 0.4553,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 16
},
{
"epoch": 0.1619047619047619,
"grad_norm": 0.6258607506752014,
"learning_rate": 0.00019974187144595432,
"loss": 0.3913,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 17
},
{
"epoch": 0.17142857142857143,
"grad_norm": 0.47939175367355347,
"learning_rate": 0.00019966289692316944,
"loss": 0.3048,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 18
},
{
"epoch": 0.18095238095238095,
"grad_norm": 0.49321815371513367,
"learning_rate": 0.00019957341762950344,
"loss": 0.3417,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 19
},
{
"epoch": 0.19047619047619047,
"grad_norm": 0.4252234101295471,
"learning_rate": 0.00019947344299634464,
"loss": 0.3033,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 20
},
{
"epoch": 0.2,
"grad_norm": 0.46225014328956604,
"learning_rate": 0.00019936298356132176,
"loss": 0.2984,
"memory/device_mem_reserved(gib)": 47.1,
"memory/max_mem_active(gib)": 45.45,
"memory/max_mem_allocated(gib)": 45.45,
"step": 21
},
{
"epoch": 0.20952380952380953,
"grad_norm": 0.4106466472148895,
"learning_rate": 0.0001992420509671936,
"loss": 0.2764,
"memory/device_mem_reserved(gib)": 47.12,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 22
},
{
"epoch": 0.21904761904761905,
"grad_norm": 0.3764067590236664,
"learning_rate": 0.00019911065796062135,
"loss": 0.2352,
"memory/device_mem_reserved(gib)": 47.12,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 23
},
{
"epoch": 0.22857142857142856,
"grad_norm": 0.38518962264060974,
"learning_rate": 0.00019896881839082556,
"loss": 0.2239,
"memory/device_mem_reserved(gib)": 47.12,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 24
},
{
"epoch": 0.23809523809523808,
"grad_norm": 0.39728403091430664,
"learning_rate": 0.00019881654720812594,
"loss": 0.2192,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 25
},
{
"epoch": 0.24761904761904763,
"grad_norm": 0.3667093813419342,
"learning_rate": 0.00019865386046236596,
"loss": 0.2834,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 26
},
{
"epoch": 0.2571428571428571,
"grad_norm": 0.32502660155296326,
"learning_rate": 0.00019848077530122083,
"loss": 0.2166,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 27
},
{
"epoch": 0.26666666666666666,
"grad_norm": 0.31225672364234924,
"learning_rate": 0.0001982973099683902,
"loss": 0.2238,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 28
},
{
"epoch": 0.2761904761904762,
"grad_norm": 0.27448731660842896,
"learning_rate": 0.00019810348380167527,
"loss": 0.1983,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 29
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.32014158368110657,
"learning_rate": 0.00019789931723094046,
"loss": 0.2554,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 30
},
{
"epoch": 0.29523809523809524,
"grad_norm": 0.39480143785476685,
"learning_rate": 0.0001976848317759601,
"loss": 0.2438,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 31
},
{
"epoch": 0.3047619047619048,
"grad_norm": 0.28004321455955505,
"learning_rate": 0.00019746005004415005,
"loss": 0.2145,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 32
},
{
"epoch": 0.3142857142857143,
"grad_norm": 0.2754361033439636,
"learning_rate": 0.00019722499572818496,
"loss": 0.2284,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.46,
"memory/max_mem_allocated(gib)": 45.46,
"step": 33
},
{
"epoch": 0.3238095238095238,
"grad_norm": 0.3160915970802307,
"learning_rate": 0.00019697969360350098,
"loss": 0.2482,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 34
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.24699093401432037,
"learning_rate": 0.00019672416952568416,
"loss": 0.1916,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 35
},
{
"epoch": 0.34285714285714286,
"grad_norm": 0.2674485743045807,
"learning_rate": 0.00019645845042774553,
"loss": 0.2204,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 36
},
{
"epoch": 0.3523809523809524,
"grad_norm": 0.33608749508857727,
"learning_rate": 0.00019618256431728194,
"loss": 0.235,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 37
},
{
"epoch": 0.3619047619047619,
"grad_norm": 0.29581594467163086,
"learning_rate": 0.00019589654027352414,
"loss": 0.2175,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 38
},
{
"epoch": 0.37142857142857144,
"grad_norm": 0.3526640832424164,
"learning_rate": 0.0001956004084442718,
"loss": 0.1989,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 39
},
{
"epoch": 0.38095238095238093,
"grad_norm": 0.2882135808467865,
"learning_rate": 0.00019529420004271567,
"loss": 0.2144,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 40
},
{
"epoch": 0.3904761904761905,
"grad_norm": 0.2675183117389679,
"learning_rate": 0.0001949779473441478,
"loss": 0.1987,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 41
},
{
"epoch": 0.4,
"grad_norm": 0.2908264696598053,
"learning_rate": 0.00019465168368255946,
"loss": 0.1893,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 42
},
{
"epoch": 0.4095238095238095,
"grad_norm": 0.3211234211921692,
"learning_rate": 0.00019431544344712776,
"loss": 0.2008,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 43
},
{
"epoch": 0.41904761904761906,
"grad_norm": 0.2795293927192688,
"learning_rate": 0.00019396926207859084,
"loss": 0.1935,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 44
},
{
"epoch": 0.42857142857142855,
"grad_norm": 0.28679871559143066,
"learning_rate": 0.00019361317606551238,
"loss": 0.1987,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 45
},
{
"epoch": 0.4380952380952381,
"grad_norm": 0.3165067732334137,
"learning_rate": 0.00019324722294043558,
"loss": 0.2118,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 46
},
{
"epoch": 0.44761904761904764,
"grad_norm": 0.25720369815826416,
"learning_rate": 0.00019287144127592704,
"loss": 0.1962,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 47
},
{
"epoch": 0.45714285714285713,
"grad_norm": 0.2577202022075653,
"learning_rate": 0.0001924858706805112,
"loss": 0.2022,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 48
},
{
"epoch": 0.4666666666666667,
"grad_norm": 0.2605520188808441,
"learning_rate": 0.0001920905517944954,
"loss": 0.1891,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 49
},
{
"epoch": 0.47619047619047616,
"grad_norm": 0.2890350818634033,
"learning_rate": 0.00019168552628568631,
"loss": 0.17,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 50
},
{
"epoch": 0.4857142857142857,
"grad_norm": 0.25532475113868713,
"learning_rate": 0.00019127083684499806,
"loss": 0.1796,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 51
},
{
"epoch": 0.49523809523809526,
"grad_norm": 0.2963904142379761,
"learning_rate": 0.00019084652718195238,
"loss": 0.19,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 52
},
{
"epoch": 0.5047619047619047,
"grad_norm": 0.24455095827579498,
"learning_rate": 0.0001904126420200716,
"loss": 0.2017,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 53
},
{
"epoch": 0.5142857142857142,
"grad_norm": 0.2518145442008972,
"learning_rate": 0.00018996922709216455,
"loss": 0.1896,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 54
},
{
"epoch": 0.5238095238095238,
"grad_norm": 0.26225805282592773,
"learning_rate": 0.00018951632913550626,
"loss": 0.1893,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 55
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.25169169902801514,
"learning_rate": 0.00018905399588691163,
"loss": 0.1939,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 56
},
{
"epoch": 0.5428571428571428,
"grad_norm": 0.23847182095050812,
"learning_rate": 0.00018858227607770398,
"loss": 0.1686,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 57
},
{
"epoch": 0.5523809523809524,
"grad_norm": 0.30888280272483826,
"learning_rate": 0.00018810121942857845,
"loss": 0.1768,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 58
},
{
"epoch": 0.5619047619047619,
"grad_norm": 0.27275580167770386,
"learning_rate": 0.00018761087664436138,
"loss": 0.1941,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 59
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.22851359844207764,
"learning_rate": 0.00018711129940866575,
"loss": 0.1658,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 60
},
{
"epoch": 0.580952380952381,
"grad_norm": 0.25515425205230713,
"learning_rate": 0.00018660254037844388,
"loss": 0.1858,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 61
},
{
"epoch": 0.5904761904761905,
"grad_norm": 0.24301765859127045,
"learning_rate": 0.00018608465317843678,
"loss": 0.1892,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 62
},
{
"epoch": 0.6,
"grad_norm": 0.24329794943332672,
"learning_rate": 0.00018555769239552233,
"loss": 0.1719,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 63
},
{
"epoch": 0.6095238095238096,
"grad_norm": 0.3206610381603241,
"learning_rate": 0.00018502171357296144,
"loss": 0.2075,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 64
},
{
"epoch": 0.6190476190476191,
"grad_norm": 0.3089618980884552,
"learning_rate": 0.00018447677320454367,
"loss": 0.185,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 65
},
{
"epoch": 0.6285714285714286,
"grad_norm": 0.291916161775589,
"learning_rate": 0.00018392292872863267,
"loss": 0.1739,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 66
},
{
"epoch": 0.638095238095238,
"grad_norm": 0.25783228874206543,
"learning_rate": 0.00018336023852211195,
"loss": 0.2061,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 67
},
{
"epoch": 0.6476190476190476,
"grad_norm": 0.310614675283432,
"learning_rate": 0.00018278876189423179,
"loss": 0.1984,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 68
},
{
"epoch": 0.6571428571428571,
"grad_norm": 0.23836325109004974,
"learning_rate": 0.00018220855908035783,
"loss": 0.1742,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 69
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.30577945709228516,
"learning_rate": 0.0001816196912356222,
"loss": 0.1954,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 70
},
{
"epoch": 0.6761904761904762,
"grad_norm": 0.3177741467952728,
"learning_rate": 0.00018102222042847737,
"loss": 0.193,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 71
},
{
"epoch": 0.6857142857142857,
"grad_norm": 0.3444816470146179,
"learning_rate": 0.00018041620963415417,
"loss": 0.1837,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 72
},
{
"epoch": 0.6952380952380952,
"grad_norm": 0.3139638900756836,
"learning_rate": 0.000179801722728024,
"loss": 0.1902,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 73
},
{
"epoch": 0.7047619047619048,
"grad_norm": 0.24754184484481812,
"learning_rate": 0.00017917882447886582,
"loss": 0.1612,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 74
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.23204036056995392,
"learning_rate": 0.00017854758054203988,
"loss": 0.1694,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 75
},
{
"epoch": 0.7238095238095238,
"grad_norm": 0.2823260426521301,
"learning_rate": 0.00017790805745256704,
"loss": 0.1739,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 76
},
{
"epoch": 0.7333333333333333,
"grad_norm": 0.3193572163581848,
"learning_rate": 0.0001772603226181159,
"loss": 0.1924,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 77
},
{
"epoch": 0.7428571428571429,
"grad_norm": 0.30530717968940735,
"learning_rate": 0.0001766044443118978,
"loss": 0.1375,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 78
},
{
"epoch": 0.7523809523809524,
"grad_norm": 0.2974378764629364,
"learning_rate": 0.00017594049166547073,
"loss": 0.1866,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 79
},
{
"epoch": 0.7619047619047619,
"grad_norm": 0.26553574204444885,
"learning_rate": 0.00017526853466145244,
"loss": 0.193,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 80
},
{
"epoch": 0.7714285714285715,
"grad_norm": 0.32059594988822937,
"learning_rate": 0.00017458864412614434,
"loss": 0.1822,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 81
},
{
"epoch": 0.780952380952381,
"grad_norm": 0.3260822594165802,
"learning_rate": 0.00017390089172206592,
"loss": 0.1734,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 82
},
{
"epoch": 0.7904761904761904,
"grad_norm": 0.2754499912261963,
"learning_rate": 0.00017320534994040148,
"loss": 0.1814,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 83
},
{
"epoch": 0.8,
"grad_norm": 0.2639872431755066,
"learning_rate": 0.00017250209209335927,
"loss": 0.1764,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 84
},
{
"epoch": 0.8095238095238095,
"grad_norm": 0.22905386984348297,
"learning_rate": 0.0001717911923064442,
"loss": 0.1365,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 85
},
{
"epoch": 0.819047619047619,
"grad_norm": 0.26005449891090393,
"learning_rate": 0.00017107272551064473,
"loss": 0.1756,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 86
},
{
"epoch": 0.8285714285714286,
"grad_norm": 0.25042417645454407,
"learning_rate": 0.00017034676743453499,
"loss": 0.15,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 87
},
{
"epoch": 0.8380952380952381,
"grad_norm": 0.34358277916908264,
"learning_rate": 0.0001696133945962927,
"loss": 0.1763,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 88
},
{
"epoch": 0.8476190476190476,
"grad_norm": 0.23855257034301758,
"learning_rate": 0.0001688726842956339,
"loss": 0.1547,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 89
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.2579350173473358,
"learning_rate": 0.0001681247146056654,
"loss": 0.1872,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 90
},
{
"epoch": 0.8666666666666667,
"grad_norm": 0.30034390091896057,
"learning_rate": 0.00016736956436465573,
"loss": 0.1558,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 91
},
{
"epoch": 0.8761904761904762,
"grad_norm": 0.28185054659843445,
"learning_rate": 0.00016660731316772505,
"loss": 0.1878,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 92
},
{
"epoch": 0.8857142857142857,
"grad_norm": 0.32514333724975586,
"learning_rate": 0.0001658380413584558,
"loss": 0.1705,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 93
},
{
"epoch": 0.8952380952380953,
"grad_norm": 0.30903860926628113,
"learning_rate": 0.0001650618300204242,
"loss": 0.1708,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 94
},
{
"epoch": 0.9047619047619048,
"grad_norm": 0.24723806977272034,
"learning_rate": 0.00016427876096865394,
"loss": 0.1571,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 95
},
{
"epoch": 0.9142857142857143,
"grad_norm": 0.2727813422679901,
"learning_rate": 0.0001634889167409923,
"loss": 0.1779,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 96
},
{
"epoch": 0.9238095238095239,
"grad_norm": 0.2521083354949951,
"learning_rate": 0.0001626923805894107,
"loss": 0.1451,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 97
},
{
"epoch": 0.9333333333333333,
"grad_norm": 0.25517475605010986,
"learning_rate": 0.00016188923647122947,
"loss": 0.1789,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 98
},
{
"epoch": 0.9428571428571428,
"grad_norm": 0.26679396629333496,
"learning_rate": 0.0001610795690402688,
"loss": 0.1835,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 99
},
{
"epoch": 0.9523809523809523,
"grad_norm": 0.22741852700710297,
"learning_rate": 0.00016026346363792567,
"loss": 0.1599,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 100
},
{
"epoch": 0.9619047619047619,
"grad_norm": 0.28691986203193665,
"learning_rate": 0.00015944100628417868,
"loss": 0.1654,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 101
},
{
"epoch": 0.9714285714285714,
"grad_norm": 0.24840302765369415,
"learning_rate": 0.00015861228366852148,
"loss": 0.1584,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 102
},
{
"epoch": 0.9809523809523809,
"grad_norm": 0.24804599583148956,
"learning_rate": 0.00015777738314082514,
"loss": 0.1391,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 103
},
{
"epoch": 0.9904761904761905,
"grad_norm": 0.2463735193014145,
"learning_rate": 0.00015693639270213136,
"loss": 0.1408,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 104
},
{
"epoch": 1.0,
"grad_norm": 0.29896217584609985,
"learning_rate": 0.000156089400995377,
"loss": 0.1666,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 105
},
{
"epoch": 1.0095238095238095,
"grad_norm": 0.2636161744594574,
"learning_rate": 0.0001552364972960506,
"loss": 0.1714,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 106
},
{
"epoch": 1.019047619047619,
"grad_norm": 0.2503117322921753,
"learning_rate": 0.00015437777150278267,
"loss": 0.1522,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 107
},
{
"epoch": 1.0285714285714285,
"grad_norm": 0.3032307028770447,
"learning_rate": 0.00015351331412787004,
"loss": 0.163,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 108
},
{
"epoch": 1.0380952380952382,
"grad_norm": 0.2652963399887085,
"learning_rate": 0.0001526432162877356,
"loss": 0.1589,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 109
},
{
"epoch": 1.0476190476190477,
"grad_norm": 0.2623184621334076,
"learning_rate": 0.00015176756969332425,
"loss": 0.1429,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 110
},
{
"epoch": 1.0571428571428572,
"grad_norm": 0.28222620487213135,
"learning_rate": 0.0001508864666404365,
"loss": 0.1469,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 111
},
{
"epoch": 1.0666666666666667,
"grad_norm": 0.28255829215049744,
"learning_rate": 0.00015000000000000001,
"loss": 0.156,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 112
},
{
"epoch": 1.0761904761904761,
"grad_norm": 0.22465772926807404,
"learning_rate": 0.00014910826320828084,
"loss": 0.131,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 113
},
{
"epoch": 1.0857142857142856,
"grad_norm": 0.30199363827705383,
"learning_rate": 0.0001482113502570349,
"loss": 0.1599,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 114
},
{
"epoch": 1.0952380952380953,
"grad_norm": 0.255288302898407,
"learning_rate": 0.00014730935568360102,
"loss": 0.1636,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 115
},
{
"epoch": 1.1047619047619048,
"grad_norm": 0.26142561435699463,
"learning_rate": 0.00014640237456093634,
"loss": 0.1518,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 116
},
{
"epoch": 1.1142857142857143,
"grad_norm": 0.30472439527511597,
"learning_rate": 0.00014549050248759547,
"loss": 0.1594,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 117
},
{
"epoch": 1.1238095238095238,
"grad_norm": 0.26955774426460266,
"learning_rate": 0.00014457383557765386,
"loss": 0.1676,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 118
},
{
"epoch": 1.1333333333333333,
"grad_norm": 0.34260597825050354,
"learning_rate": 0.00014365247045057734,
"loss": 0.1583,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 119
},
{
"epoch": 1.1428571428571428,
"grad_norm": 0.3136855661869049,
"learning_rate": 0.0001427265042210381,
"loss": 0.1558,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 120
},
{
"epoch": 1.1523809523809523,
"grad_norm": 0.27992814779281616,
"learning_rate": 0.00014179603448867835,
"loss": 0.1509,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 121
},
{
"epoch": 1.161904761904762,
"grad_norm": 0.34893789887428284,
"learning_rate": 0.00014086115932782314,
"loss": 0.1395,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 122
},
{
"epoch": 1.1714285714285715,
"grad_norm": 0.3308560252189636,
"learning_rate": 0.0001399219772771431,
"loss": 0.1591,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 123
},
{
"epoch": 1.180952380952381,
"grad_norm": 0.2736824154853821,
"learning_rate": 0.00013897858732926793,
"loss": 0.1604,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 124
},
{
"epoch": 1.1904761904761905,
"grad_norm": 0.28148430585861206,
"learning_rate": 0.0001380310889203526,
"loss": 0.1368,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 125
},
{
"epoch": 1.2,
"grad_norm": 0.25899434089660645,
"learning_rate": 0.00013707958191959608,
"loss": 0.15,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 126
},
{
"epoch": 1.2095238095238094,
"grad_norm": 0.32127872109413147,
"learning_rate": 0.00013612416661871533,
"loss": 0.1522,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 127
},
{
"epoch": 1.2190476190476192,
"grad_norm": 0.350427508354187,
"learning_rate": 0.00013516494372137368,
"loss": 0.1361,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 128
},
{
"epoch": 1.2285714285714286,
"grad_norm": 0.36603009700775146,
"learning_rate": 0.00013420201433256689,
"loss": 0.1527,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 129
},
{
"epoch": 1.2380952380952381,
"grad_norm": 0.2870017886161804,
"learning_rate": 0.00013323547994796597,
"loss": 0.152,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 130
},
{
"epoch": 1.2476190476190476,
"grad_norm": 0.29138877987861633,
"learning_rate": 0.0001322654424432195,
"loss": 0.1599,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 131
},
{
"epoch": 1.2571428571428571,
"grad_norm": 0.3358840048313141,
"learning_rate": 0.00013129200406321545,
"loss": 0.1668,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 132
},
{
"epoch": 1.2666666666666666,
"grad_norm": 0.3152446746826172,
"learning_rate": 0.00013031526741130435,
"loss": 0.1382,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 133
},
{
"epoch": 1.276190476190476,
"grad_norm": 0.2602708041667938,
"learning_rate": 0.00012933533543848461,
"loss": 0.1441,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 134
},
{
"epoch": 1.2857142857142856,
"grad_norm": 0.27083972096443176,
"learning_rate": 0.0001283523114325511,
"loss": 0.1517,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 135
},
{
"epoch": 1.2952380952380953,
"grad_norm": 0.27242034673690796,
"learning_rate": 0.0001273662990072083,
"loss": 0.1511,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 136
},
{
"epoch": 1.3047619047619048,
"grad_norm": 0.28473100066185,
"learning_rate": 0.0001263774020911492,
"loss": 0.1339,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 137
},
{
"epoch": 1.3142857142857143,
"grad_norm": 0.27939581871032715,
"learning_rate": 0.0001253857249171008,
"loss": 0.1414,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 138
},
{
"epoch": 1.3238095238095238,
"grad_norm": 0.2833685874938965,
"learning_rate": 0.00012439137201083773,
"loss": 0.139,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 139
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.3193693161010742,
"learning_rate": 0.0001233944481801649,
"loss": 0.159,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 140
},
{
"epoch": 1.342857142857143,
"grad_norm": 0.28556281328201294,
"learning_rate": 0.0001223950585038703,
"loss": 0.1549,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 141
},
{
"epoch": 1.3523809523809525,
"grad_norm": 0.2670241594314575,
"learning_rate": 0.00012139330832064974,
"loss": 0.1595,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 142
},
{
"epoch": 1.361904761904762,
"grad_norm": 0.20607773959636688,
"learning_rate": 0.00012038930321800346,
"loss": 0.1299,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 143
},
{
"epoch": 1.3714285714285714,
"grad_norm": 0.2973734736442566,
"learning_rate": 0.00011938314902110701,
"loss": 0.1552,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 144
},
{
"epoch": 1.380952380952381,
"grad_norm": 0.3032817244529724,
"learning_rate": 0.00011837495178165706,
"loss": 0.1554,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 145
},
{
"epoch": 1.3904761904761904,
"grad_norm": 0.2765556871891022,
"learning_rate": 0.00011736481776669306,
"loss": 0.1346,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 146
},
{
"epoch": 1.4,
"grad_norm": 0.24869301915168762,
"learning_rate": 0.00011635285344739651,
"loss": 0.1573,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 147
},
{
"epoch": 1.4095238095238094,
"grad_norm": 0.28329774737358093,
"learning_rate": 0.00011533916548786857,
"loss": 0.166,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 148
},
{
"epoch": 1.4190476190476191,
"grad_norm": 0.31719744205474854,
"learning_rate": 0.00011432386073388717,
"loss": 0.1556,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 149
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.27932024002075195,
"learning_rate": 0.00011330704620164538,
"loss": 0.1722,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 150
},
{
"epoch": 1.438095238095238,
"grad_norm": 0.2447134554386139,
"learning_rate": 0.00011228882906647142,
"loss": 0.1318,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 151
},
{
"epoch": 1.4476190476190476,
"grad_norm": 0.2684175670146942,
"learning_rate": 0.00011126931665153212,
"loss": 0.1465,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 152
},
{
"epoch": 1.457142857142857,
"grad_norm": 0.24706321954727173,
"learning_rate": 0.00011024861641652071,
"loss": 0.1226,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 153
},
{
"epoch": 1.4666666666666668,
"grad_norm": 0.2958153486251831,
"learning_rate": 0.00010922683594633021,
"loss": 0.1447,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 154
},
{
"epoch": 1.4761904761904763,
"grad_norm": 0.305876761674881,
"learning_rate": 0.00010820408293971378,
"loss": 0.1566,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 155
},
{
"epoch": 1.4857142857142858,
"grad_norm": 0.2755098342895508,
"learning_rate": 0.00010718046519793276,
"loss": 0.155,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 156
},
{
"epoch": 1.4952380952380953,
"grad_norm": 0.2550102174282074,
"learning_rate": 0.00010615609061339432,
"loss": 0.1227,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 157
},
{
"epoch": 1.5047619047619047,
"grad_norm": 0.26932892203330994,
"learning_rate": 0.00010513106715827896,
"loss": 0.1533,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 158
},
{
"epoch": 1.5142857142857142,
"grad_norm": 0.24724788963794708,
"learning_rate": 0.00010410550287316034,
"loss": 0.1346,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 159
},
{
"epoch": 1.5238095238095237,
"grad_norm": 0.2605031132698059,
"learning_rate": 0.00010307950585561706,
"loss": 0.1319,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 160
},
{
"epoch": 1.5333333333333332,
"grad_norm": 0.28353723883628845,
"learning_rate": 0.00010205318424883905,
"loss": 0.1386,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 161
},
{
"epoch": 1.5428571428571427,
"grad_norm": 0.31240761280059814,
"learning_rate": 0.00010102664623022899,
"loss": 0.1468,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 162
},
{
"epoch": 1.5523809523809524,
"grad_norm": 0.2614494562149048,
"learning_rate": 0.0001,
"loss": 0.146,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.47,
"memory/max_mem_allocated(gib)": 45.47,
"step": 163
},
{
"epoch": 1.561904761904762,
"grad_norm": 0.28343483805656433,
"learning_rate": 9.897335376977102e-05,
"loss": 0.1531,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 164
},
{
"epoch": 1.5714285714285714,
"grad_norm": 0.2741709053516388,
"learning_rate": 9.794681575116097e-05,
"loss": 0.1304,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 165
},
{
"epoch": 1.580952380952381,
"grad_norm": 0.326215922832489,
"learning_rate": 9.692049414438299e-05,
"loss": 0.1551,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 166
},
{
"epoch": 1.5904761904761906,
"grad_norm": 0.3120001256465912,
"learning_rate": 9.58944971268397e-05,
"loss": 0.1387,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 167
},
{
"epoch": 1.6,
"grad_norm": 0.2673357129096985,
"learning_rate": 9.486893284172102e-05,
"loss": 0.1389,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 168
},
{
"epoch": 1.6095238095238096,
"grad_norm": 0.3184024691581726,
"learning_rate": 9.384390938660572e-05,
"loss": 0.1388,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 169
},
{
"epoch": 1.619047619047619,
"grad_norm": 0.28744614124298096,
"learning_rate": 9.281953480206725e-05,
"loss": 0.1246,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 170
},
{
"epoch": 1.6285714285714286,
"grad_norm": 0.28229352831840515,
"learning_rate": 9.179591706028626e-05,
"loss": 0.1637,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 171
},
{
"epoch": 1.638095238095238,
"grad_norm": 0.2750794589519501,
"learning_rate": 9.077316405366981e-05,
"loss": 0.1176,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 172
},
{
"epoch": 1.6476190476190475,
"grad_norm": 0.28753307461738586,
"learning_rate": 8.975138358347931e-05,
"loss": 0.1517,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 173
},
{
"epoch": 1.657142857142857,
"grad_norm": 0.2737794518470764,
"learning_rate": 8.87306833484679e-05,
"loss": 0.1604,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 174
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.2456161379814148,
"learning_rate": 8.77111709335286e-05,
"loss": 0.1441,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 175
},
{
"epoch": 1.6761904761904762,
"grad_norm": 0.2895718812942505,
"learning_rate": 8.669295379835467e-05,
"loss": 0.152,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 176
},
{
"epoch": 1.6857142857142857,
"grad_norm": 0.3128141462802887,
"learning_rate": 8.567613926611288e-05,
"loss": 0.1692,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 177
},
{
"epoch": 1.6952380952380952,
"grad_norm": 0.29811859130859375,
"learning_rate": 8.466083451213144e-05,
"loss": 0.1351,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 178
},
{
"epoch": 1.704761904761905,
"grad_norm": 0.2686918377876282,
"learning_rate": 8.364714655260349e-05,
"loss": 0.1318,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 179
},
{
"epoch": 1.7142857142857144,
"grad_norm": 0.31242185831069946,
"learning_rate": 8.263518223330697e-05,
"loss": 0.1512,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 180
},
{
"epoch": 1.723809523809524,
"grad_norm": 0.3040373623371124,
"learning_rate": 8.162504821834295e-05,
"loss": 0.1464,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 181
},
{
"epoch": 1.7333333333333334,
"grad_norm": 0.2847082018852234,
"learning_rate": 8.0616850978893e-05,
"loss": 0.1586,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 182
},
{
"epoch": 1.7428571428571429,
"grad_norm": 0.27451273798942566,
"learning_rate": 7.961069678199658e-05,
"loss": 0.1553,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 183
},
{
"epoch": 1.7523809523809524,
"grad_norm": 0.3282047212123871,
"learning_rate": 7.860669167935028e-05,
"loss": 0.1559,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 184
},
{
"epoch": 1.7619047619047619,
"grad_norm": 0.2895129919052124,
"learning_rate": 7.760494149612971e-05,
"loss": 0.1431,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 185
},
{
"epoch": 1.7714285714285714,
"grad_norm": 0.2585572898387909,
"learning_rate": 7.660555181983518e-05,
"loss": 0.1274,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 186
},
{
"epoch": 1.7809523809523808,
"grad_norm": 0.2550938129425049,
"learning_rate": 7.560862798916228e-05,
"loss": 0.1482,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 187
},
{
"epoch": 1.7904761904761903,
"grad_norm": 0.267459511756897,
"learning_rate": 7.461427508289922e-05,
"loss": 0.1425,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 188
},
{
"epoch": 1.8,
"grad_norm": 0.273129940032959,
"learning_rate": 7.36225979088508e-05,
"loss": 0.118,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 189
},
{
"epoch": 1.8095238095238095,
"grad_norm": 0.33791351318359375,
"learning_rate": 7.263370099279172e-05,
"loss": 0.146,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 190
},
{
"epoch": 1.819047619047619,
"grad_norm": 0.3014010488986969,
"learning_rate": 7.164768856744892e-05,
"loss": 0.1461,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 191
},
{
"epoch": 1.8285714285714287,
"grad_norm": 0.32012397050857544,
"learning_rate": 7.066466456151541e-05,
"loss": 0.1782,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 192
},
{
"epoch": 1.8380952380952382,
"grad_norm": 0.25507038831710815,
"learning_rate": 6.968473258869566e-05,
"loss": 0.1319,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 193
},
{
"epoch": 1.8476190476190477,
"grad_norm": 0.28470760583877563,
"learning_rate": 6.870799593678459e-05,
"loss": 0.159,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 194
},
{
"epoch": 1.8571428571428572,
"grad_norm": 0.25228601694107056,
"learning_rate": 6.773455755678054e-05,
"loss": 0.1284,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 195
},
{
"epoch": 1.8666666666666667,
"grad_norm": 0.28710344433784485,
"learning_rate": 6.676452005203406e-05,
"loss": 0.1338,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 196
},
{
"epoch": 1.8761904761904762,
"grad_norm": 0.29423221945762634,
"learning_rate": 6.579798566743314e-05,
"loss": 0.1335,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 197
},
{
"epoch": 1.8857142857142857,
"grad_norm": 0.31776273250579834,
"learning_rate": 6.483505627862632e-05,
"loss": 0.1509,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 198
},
{
"epoch": 1.8952380952380952,
"grad_norm": 0.2739746868610382,
"learning_rate": 6.387583338128471e-05,
"loss": 0.1195,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 199
},
{
"epoch": 1.9047619047619047,
"grad_norm": 0.28000062704086304,
"learning_rate": 6.292041808040393e-05,
"loss": 0.1444,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 200
},
{
"epoch": 1.9142857142857141,
"grad_norm": 0.2637692987918854,
"learning_rate": 6.196891107964744e-05,
"loss": 0.1412,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 201
},
{
"epoch": 1.9238095238095239,
"grad_norm": 0.2977365255355835,
"learning_rate": 6.102141267073207e-05,
"loss": 0.1224,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 202
},
{
"epoch": 1.9333333333333333,
"grad_norm": 0.2656545042991638,
"learning_rate": 6.007802272285693e-05,
"loss": 0.1235,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 203
},
{
"epoch": 1.9428571428571428,
"grad_norm": 0.3028334975242615,
"learning_rate": 5.913884067217685e-05,
"loss": 0.1544,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 204
},
{
"epoch": 1.9523809523809523,
"grad_norm": 0.32587432861328125,
"learning_rate": 5.82039655113217e-05,
"loss": 0.1484,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 205
},
{
"epoch": 1.961904761904762,
"grad_norm": 0.2999066710472107,
"learning_rate": 5.727349577896194e-05,
"loss": 0.1392,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 206
},
{
"epoch": 1.9714285714285715,
"grad_norm": 0.2822760045528412,
"learning_rate": 5.634752954942264e-05,
"loss": 0.1334,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 207
},
{
"epoch": 1.980952380952381,
"grad_norm": 0.4100743234157562,
"learning_rate": 5.542616442234618e-05,
"loss": 0.1295,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 208
},
{
"epoch": 1.9904761904761905,
"grad_norm": 0.22527122497558594,
"learning_rate": 5.450949751240456e-05,
"loss": 0.1377,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 209
},
{
"epoch": 2.0,
"grad_norm": 0.33393946290016174,
"learning_rate": 5.359762543906368e-05,
"loss": 0.1251,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 210
},
{
"epoch": 2.0095238095238095,
"grad_norm": 0.2823657989501953,
"learning_rate": 5.269064431639901e-05,
"loss": 0.1443,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 211
},
{
"epoch": 2.019047619047619,
"grad_norm": 0.277386873960495,
"learning_rate": 5.178864974296511e-05,
"loss": 0.1333,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 212
},
{
"epoch": 2.0285714285714285,
"grad_norm": 0.2936922311782837,
"learning_rate": 5.0891736791719213e-05,
"loss": 0.1209,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 213
},
{
"epoch": 2.038095238095238,
"grad_norm": 0.33104684948921204,
"learning_rate": 5.000000000000002e-05,
"loss": 0.1344,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 214
},
{
"epoch": 2.0476190476190474,
"grad_norm": 0.2811061143875122,
"learning_rate": 4.911353335956352e-05,
"loss": 0.1182,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 215
},
{
"epoch": 2.057142857142857,
"grad_norm": 0.3159407377243042,
"learning_rate": 4.823243030667576e-05,
"loss": 0.1506,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 216
},
{
"epoch": 2.066666666666667,
"grad_norm": 0.2573259472846985,
"learning_rate": 4.735678371226441e-05,
"loss": 0.115,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 217
},
{
"epoch": 2.0761904761904764,
"grad_norm": 0.2847664952278137,
"learning_rate": 4.648668587212997e-05,
"loss": 0.1596,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 218
},
{
"epoch": 2.085714285714286,
"grad_norm": 0.31137990951538086,
"learning_rate": 4.562222849721735e-05,
"loss": 0.1389,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 219
},
{
"epoch": 2.0952380952380953,
"grad_norm": 0.29724523425102234,
"learning_rate": 4.476350270394942e-05,
"loss": 0.1286,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 220
},
{
"epoch": 2.104761904761905,
"grad_norm": 0.3148998022079468,
"learning_rate": 4.391059900462304e-05,
"loss": 0.1311,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 221
},
{
"epoch": 2.1142857142857143,
"grad_norm": 0.29536697268486023,
"learning_rate": 4.306360729786867e-05,
"loss": 0.1382,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 222
},
{
"epoch": 2.123809523809524,
"grad_norm": 0.29875388741493225,
"learning_rate": 4.222261685917489e-05,
"loss": 0.1321,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 223
},
{
"epoch": 2.1333333333333333,
"grad_norm": 0.27502039074897766,
"learning_rate": 4.1387716331478565e-05,
"loss": 0.1137,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.48,
"memory/max_mem_allocated(gib)": 45.48,
"step": 224
},
{
"epoch": 2.142857142857143,
"grad_norm": 0.3302501142024994,
"learning_rate": 4.055899371582133e-05,
"loss": 0.145,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 225
},
{
"epoch": 2.1523809523809523,
"grad_norm": 0.31309428811073303,
"learning_rate": 3.973653636207437e-05,
"loss": 0.1336,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 226
},
{
"epoch": 2.1619047619047618,
"grad_norm": 0.3851131796836853,
"learning_rate": 3.8920430959731226e-05,
"loss": 0.128,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 227
},
{
"epoch": 2.1714285714285713,
"grad_norm": 0.30255237221717834,
"learning_rate": 3.811076352877054e-05,
"loss": 0.1144,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 228
},
{
"epoch": 2.1809523809523808,
"grad_norm": 0.30191734433174133,
"learning_rate": 3.7307619410589376e-05,
"loss": 0.1177,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 229
},
{
"epoch": 2.1904761904761907,
"grad_norm": 0.256660521030426,
"learning_rate": 3.651108325900773e-05,
"loss": 0.1013,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 230
},
{
"epoch": 2.2,
"grad_norm": 0.26110267639160156,
"learning_rate": 3.5721239031346066e-05,
"loss": 0.1307,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 231
},
{
"epoch": 2.2095238095238097,
"grad_norm": 0.313667893409729,
"learning_rate": 3.493816997957582e-05,
"loss": 0.1524,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 232
},
{
"epoch": 2.219047619047619,
"grad_norm": 0.3256130516529083,
"learning_rate": 3.416195864154425e-05,
"loss": 0.1375,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 233
},
{
"epoch": 2.2285714285714286,
"grad_norm": 0.5965404510498047,
"learning_rate": 3.339268683227499e-05,
"loss": 0.1346,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 234
},
{
"epoch": 2.238095238095238,
"grad_norm": 0.25208190083503723,
"learning_rate": 3.263043563534428e-05,
"loss": 0.1068,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 235
},
{
"epoch": 2.2476190476190476,
"grad_norm": 0.24084089696407318,
"learning_rate": 3.187528539433458e-05,
"loss": 0.1093,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 236
},
{
"epoch": 2.257142857142857,
"grad_norm": 0.2939463257789612,
"learning_rate": 3.112731570436614e-05,
"loss": 0.1348,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 237
},
{
"epoch": 2.2666666666666666,
"grad_norm": 0.24885237216949463,
"learning_rate": 3.0386605403707346e-05,
"loss": 0.1158,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 238
},
{
"epoch": 2.276190476190476,
"grad_norm": 0.2684744894504547,
"learning_rate": 2.9653232565465016e-05,
"loss": 0.1094,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 239
},
{
"epoch": 2.2857142857142856,
"grad_norm": 0.33237671852111816,
"learning_rate": 2.8927274489355293e-05,
"loss": 0.1227,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 240
},
{
"epoch": 2.295238095238095,
"grad_norm": 0.2611919641494751,
"learning_rate": 2.8208807693555818e-05,
"loss": 0.1,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 241
},
{
"epoch": 2.3047619047619046,
"grad_norm": 0.34490758180618286,
"learning_rate": 2.7497907906640742e-05,
"loss": 0.124,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 242
},
{
"epoch": 2.314285714285714,
"grad_norm": 0.27448931336402893,
"learning_rate": 2.679465005959856e-05,
"loss": 0.1277,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 243
},
{
"epoch": 2.323809523809524,
"grad_norm": 0.2973083257675171,
"learning_rate": 2.6099108277934103e-05,
"loss": 0.1344,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 244
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.30860233306884766,
"learning_rate": 2.541135587385568e-05,
"loss": 0.1308,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 245
},
{
"epoch": 2.342857142857143,
"grad_norm": 0.26857587695121765,
"learning_rate": 2.4731465338547556e-05,
"loss": 0.1386,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 246
},
{
"epoch": 2.3523809523809525,
"grad_norm": 0.28812259435653687,
"learning_rate": 2.405950833452928e-05,
"loss": 0.1175,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 247
},
{
"epoch": 2.361904761904762,
"grad_norm": 0.2757267951965332,
"learning_rate": 2.339555568810221e-05,
"loss": 0.1144,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 248
},
{
"epoch": 2.3714285714285714,
"grad_norm": 0.26114368438720703,
"learning_rate": 2.2739677381884115e-05,
"loss": 0.1195,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 249
},
{
"epoch": 2.380952380952381,
"grad_norm": 0.2587968707084656,
"learning_rate": 2.2091942547432955e-05,
"loss": 0.1395,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 250
},
{
"epoch": 2.3904761904761904,
"grad_norm": 0.33188945055007935,
"learning_rate": 2.1452419457960137e-05,
"loss": 0.1456,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 251
},
{
"epoch": 2.4,
"grad_norm": 0.29628974199295044,
"learning_rate": 2.0821175521134207e-05,
"loss": 0.1236,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 252
},
{
"epoch": 2.4095238095238094,
"grad_norm": 0.28231939673423767,
"learning_rate": 2.0198277271976052e-05,
"loss": 0.1265,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 253
},
{
"epoch": 2.419047619047619,
"grad_norm": 0.3047321140766144,
"learning_rate": 1.9583790365845822e-05,
"loss": 0.0983,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 254
},
{
"epoch": 2.4285714285714284,
"grad_norm": 0.29782846570014954,
"learning_rate": 1.8977779571522646e-05,
"loss": 0.1355,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 255
},
{
"epoch": 2.4380952380952383,
"grad_norm": 0.29758942127227783,
"learning_rate": 1.8380308764377842e-05,
"loss": 0.1258,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 256
},
{
"epoch": 2.447619047619048,
"grad_norm": 0.35419225692749023,
"learning_rate": 1.7791440919642177e-05,
"loss": 0.1163,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 257
},
{
"epoch": 2.4571428571428573,
"grad_norm": 0.2910211682319641,
"learning_rate": 1.7211238105768214e-05,
"loss": 0.1312,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 258
},
{
"epoch": 2.466666666666667,
"grad_norm": 0.27445051074028015,
"learning_rate": 1.663976147788806e-05,
"loss": 0.1384,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 259
},
{
"epoch": 2.4761904761904763,
"grad_norm": 0.3003697097301483,
"learning_rate": 1.607707127136734e-05,
"loss": 0.1288,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 260
},
{
"epoch": 2.4857142857142858,
"grad_norm": 0.2700045108795166,
"learning_rate": 1.5523226795456347e-05,
"loss": 0.1086,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 261
},
{
"epoch": 2.4952380952380953,
"grad_norm": 0.30866676568984985,
"learning_rate": 1.4978286427038601e-05,
"loss": 0.144,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 262
},
{
"epoch": 2.5047619047619047,
"grad_norm": 0.29147791862487793,
"learning_rate": 1.444230760447769e-05,
"loss": 0.1129,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 263
},
{
"epoch": 2.5142857142857142,
"grad_norm": 0.29668715596199036,
"learning_rate": 1.3915346821563235e-05,
"loss": 0.1346,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 264
},
{
"epoch": 2.5238095238095237,
"grad_norm": 0.2867170572280884,
"learning_rate": 1.339745962155613e-05,
"loss": 0.1216,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 265
},
{
"epoch": 2.533333333333333,
"grad_norm": 0.3114704489707947,
"learning_rate": 1.2888700591334223e-05,
"loss": 0.1229,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 266
},
{
"epoch": 2.5428571428571427,
"grad_norm": 0.30811193585395813,
"learning_rate": 1.2389123355638654e-05,
"loss": 0.1441,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 267
},
{
"epoch": 2.552380952380952,
"grad_norm": 0.3116905391216278,
"learning_rate": 1.1898780571421552e-05,
"loss": 0.1287,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 268
},
{
"epoch": 2.5619047619047617,
"grad_norm": 0.28540119528770447,
"learning_rate": 1.141772392229601e-05,
"loss": 0.1263,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 269
},
{
"epoch": 2.571428571428571,
"grad_norm": 0.4764971137046814,
"learning_rate": 1.0946004113088381e-05,
"loss": 0.1104,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 270
},
{
"epoch": 2.580952380952381,
"grad_norm": 0.3028562664985657,
"learning_rate": 1.0483670864493778e-05,
"loss": 0.1447,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 271
},
{
"epoch": 2.5904761904761906,
"grad_norm": 0.2907711863517761,
"learning_rate": 1.0030772907835483e-05,
"loss": 0.1231,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 272
},
{
"epoch": 2.6,
"grad_norm": 0.2883255183696747,
"learning_rate": 9.587357979928413e-06,
"loss": 0.1295,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 273
},
{
"epoch": 2.6095238095238096,
"grad_norm": 0.2747456729412079,
"learning_rate": 9.153472818047625e-06,
"loss": 0.1436,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 274
},
{
"epoch": 2.619047619047619,
"grad_norm": 0.2808170020580292,
"learning_rate": 8.729163155001974e-06,
"loss": 0.1517,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 275
},
{
"epoch": 2.6285714285714286,
"grad_norm": 0.25573983788490295,
"learning_rate": 8.314473714313719e-06,
"loss": 0.1178,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 276
},
{
"epoch": 2.638095238095238,
"grad_norm": 0.3509344756603241,
"learning_rate": 7.909448205504632e-06,
"loss": 0.1356,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 277
},
{
"epoch": 2.6476190476190475,
"grad_norm": 0.24175764620304108,
"learning_rate": 7.514129319488839e-06,
"loss": 0.1017,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.49,
"memory/max_mem_allocated(gib)": 45.49,
"step": 278
},
{
"epoch": 2.657142857142857,
"grad_norm": 0.29153692722320557,
"learning_rate": 7.128558724072976e-06,
"loss": 0.1202,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 279
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.2581385672092438,
"learning_rate": 6.75277705956443e-06,
"loss": 0.1148,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 280
},
{
"epoch": 2.6761904761904765,
"grad_norm": 0.2992369532585144,
"learning_rate": 6.386823934487618e-06,
"loss": 0.1408,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 281
},
{
"epoch": 2.685714285714286,
"grad_norm": 0.26200878620147705,
"learning_rate": 6.030737921409169e-06,
"loss": 0.1104,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 282
},
{
"epoch": 2.6952380952380954,
"grad_norm": 0.3082549571990967,
"learning_rate": 5.684556552872256e-06,
"loss": 0.1312,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 283
},
{
"epoch": 2.704761904761905,
"grad_norm": 0.27713072299957275,
"learning_rate": 5.348316317440549e-06,
"loss": 0.1184,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 284
},
{
"epoch": 2.7142857142857144,
"grad_norm": 0.3070668876171112,
"learning_rate": 5.0220526558522274e-06,
"loss": 0.1323,
"memory/device_mem_reserved(gib)": 47.13,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 285
},
{
"epoch": 2.723809523809524,
"grad_norm": 0.2855824828147888,
"learning_rate": 4.705799957284351e-06,
"loss": 0.1352,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 286
},
{
"epoch": 2.7333333333333334,
"grad_norm": 0.2835365831851959,
"learning_rate": 4.399591555728233e-06,
"loss": 0.1111,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 287
},
{
"epoch": 2.742857142857143,
"grad_norm": 0.23606140911579132,
"learning_rate": 4.103459726475889e-06,
"loss": 0.1099,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 288
},
{
"epoch": 2.7523809523809524,
"grad_norm": 0.32546883821487427,
"learning_rate": 3.817435682718096e-06,
"loss": 0.1271,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 289
},
{
"epoch": 2.761904761904762,
"grad_norm": 0.29360413551330566,
"learning_rate": 3.541549572254488e-06,
"loss": 0.1328,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 290
},
{
"epoch": 2.7714285714285714,
"grad_norm": 0.2728084623813629,
"learning_rate": 3.275830474315855e-06,
"loss": 0.1279,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 291
},
{
"epoch": 2.780952380952381,
"grad_norm": 0.2540973424911499,
"learning_rate": 3.0203063964990617e-06,
"loss": 0.1172,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 292
},
{
"epoch": 2.7904761904761903,
"grad_norm": 0.2903876304626465,
"learning_rate": 2.7750042718150516e-06,
"loss": 0.1254,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 293
},
{
"epoch": 2.8,
"grad_norm": 0.29096516966819763,
"learning_rate": 2.539949955849985e-06,
"loss": 0.1338,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 294
},
{
"epoch": 2.8095238095238093,
"grad_norm": 0.27256572246551514,
"learning_rate": 2.315168224039932e-06,
"loss": 0.1432,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 295
},
{
"epoch": 2.819047619047619,
"grad_norm": 0.3155645728111267,
"learning_rate": 2.100682769059548e-06,
"loss": 0.1174,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 296
},
{
"epoch": 2.8285714285714287,
"grad_norm": 0.3359803259372711,
"learning_rate": 1.8965161983247493e-06,
"loss": 0.1325,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 297
},
{
"epoch": 2.8380952380952382,
"grad_norm": 0.30490148067474365,
"learning_rate": 1.7026900316098215e-06,
"loss": 0.1162,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 298
},
{
"epoch": 2.8476190476190477,
"grad_norm": 0.29144126176834106,
"learning_rate": 1.5192246987791981e-06,
"loss": 0.1226,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 299
},
{
"epoch": 2.857142857142857,
"grad_norm": 0.2892225682735443,
"learning_rate": 1.3461395376340502e-06,
"loss": 0.1336,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 300
},
{
"epoch": 2.8666666666666667,
"grad_norm": 0.27009886503219604,
"learning_rate": 1.1834527918740623e-06,
"loss": 0.1092,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 301
},
{
"epoch": 2.876190476190476,
"grad_norm": 0.2559907138347626,
"learning_rate": 1.0311816091744698e-06,
"loss": 0.1192,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 302
},
{
"epoch": 2.8857142857142857,
"grad_norm": 0.3120160400867462,
"learning_rate": 8.893420393786489e-07,
"loss": 0.1324,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 303
},
{
"epoch": 2.895238095238095,
"grad_norm": 0.3137604892253876,
"learning_rate": 7.579490328064265e-07,
"loss": 0.1284,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 304
},
{
"epoch": 2.9047619047619047,
"grad_norm": 0.27408653497695923,
"learning_rate": 6.370164386782285e-07,
"loss": 0.122,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 305
},
{
"epoch": 2.914285714285714,
"grad_norm": 0.3108055293560028,
"learning_rate": 5.265570036553813e-07,
"loss": 0.1196,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 306
},
{
"epoch": 2.923809523809524,
"grad_norm": 0.2910846471786499,
"learning_rate": 4.2658237049655323e-07,
"loss": 0.1293,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 307
},
{
"epoch": 2.9333333333333336,
"grad_norm": 0.2849004566669464,
"learning_rate": 3.371030768305583e-07,
"loss": 0.1333,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 308
},
{
"epoch": 2.942857142857143,
"grad_norm": 0.30985507369041443,
"learning_rate": 2.5812855404568905e-07,
"loss": 0.1446,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 309
},
{
"epoch": 2.9523809523809526,
"grad_norm": 0.30461379885673523,
"learning_rate": 1.8966712629558957e-07,
"loss": 0.1203,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 310
},
{
"epoch": 2.961904761904762,
"grad_norm": 0.29812026023864746,
"learning_rate": 1.3172600962190197e-07,
"loss": 0.1275,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 311
},
{
"epoch": 2.9714285714285715,
"grad_norm": 0.29495131969451904,
"learning_rate": 8.43113111936189e-08,
"loss": 0.1212,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 312
},
{
"epoch": 2.980952380952381,
"grad_norm": 0.27057307958602905,
"learning_rate": 4.74280286634099e-08,
"loss": 0.1146,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 313
},
{
"epoch": 2.9904761904761905,
"grad_norm": 0.3303758203983307,
"learning_rate": 2.108004964086474e-08,
"loss": 0.1273,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 314
},
{
"epoch": 3.0,
"grad_norm": 0.26469886302948,
"learning_rate": 5.270151282688041e-09,
"loss": 0.1281,
"memory/device_mem_reserved(gib)": 47.14,
"memory/max_mem_active(gib)": 45.5,
"memory/max_mem_allocated(gib)": 45.5,
"step": 315
}
],
"logging_steps": 1,
"max_steps": 315,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.864148423206502e+16,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}