pphuc25's picture
End of training
a39c36e
raw
history blame contribute delete
No virus
80.9 kB
{
"best_metric": 0.15402325987815857,
"best_model_checkpoint": "whisper-small-full-data-language-v2-20ep/checkpoint-60000",
"epoch": 20.003133322888925,
"global_step": 63840,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 2.5000000000000004e-07,
"loss": 2.0939,
"step": 100
},
{
"epoch": 0.06,
"learning_rate": 5.000000000000001e-07,
"loss": 1.2605,
"step": 200
},
{
"epoch": 0.09,
"learning_rate": 7.5e-07,
"loss": 0.5371,
"step": 300
},
{
"epoch": 0.13,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.3785,
"step": 400
},
{
"epoch": 0.16,
"learning_rate": 1.25e-06,
"loss": 0.3296,
"step": 500
},
{
"epoch": 0.19,
"learning_rate": 1.5e-06,
"loss": 0.3046,
"step": 600
},
{
"epoch": 0.22,
"learning_rate": 1.7500000000000002e-06,
"loss": 0.2878,
"step": 700
},
{
"epoch": 0.25,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.2725,
"step": 800
},
{
"epoch": 0.28,
"learning_rate": 2.25e-06,
"loss": 0.2628,
"step": 900
},
{
"epoch": 0.31,
"learning_rate": 2.5e-06,
"loss": 0.2515,
"step": 1000
},
{
"epoch": 0.34,
"learning_rate": 2.7500000000000004e-06,
"loss": 0.2439,
"step": 1100
},
{
"epoch": 0.38,
"learning_rate": 3e-06,
"loss": 0.2333,
"step": 1200
},
{
"epoch": 0.41,
"learning_rate": 3.2500000000000002e-06,
"loss": 0.2262,
"step": 1300
},
{
"epoch": 0.44,
"learning_rate": 3.5000000000000004e-06,
"loss": 0.222,
"step": 1400
},
{
"epoch": 0.47,
"learning_rate": 3.75e-06,
"loss": 0.2162,
"step": 1500
},
{
"epoch": 0.5,
"learning_rate": 4.000000000000001e-06,
"loss": 0.2122,
"step": 1600
},
{
"epoch": 0.53,
"learning_rate": 4.250000000000001e-06,
"loss": 0.2074,
"step": 1700
},
{
"epoch": 0.56,
"learning_rate": 4.5e-06,
"loss": 0.2034,
"step": 1800
},
{
"epoch": 0.6,
"learning_rate": 4.75e-06,
"loss": 0.1998,
"step": 1900
},
{
"epoch": 0.63,
"learning_rate": 5e-06,
"loss": 0.1961,
"step": 2000
},
{
"epoch": 0.66,
"learning_rate": 5.25e-06,
"loss": 0.1892,
"step": 2100
},
{
"epoch": 0.69,
"learning_rate": 5.500000000000001e-06,
"loss": 0.1873,
"step": 2200
},
{
"epoch": 0.72,
"learning_rate": 5.750000000000001e-06,
"loss": 0.1843,
"step": 2300
},
{
"epoch": 0.75,
"learning_rate": 6e-06,
"loss": 0.1881,
"step": 2400
},
{
"epoch": 0.78,
"learning_rate": 6.25e-06,
"loss": 0.1813,
"step": 2500
},
{
"epoch": 0.81,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.1828,
"step": 2600
},
{
"epoch": 0.85,
"learning_rate": 6.750000000000001e-06,
"loss": 0.1782,
"step": 2700
},
{
"epoch": 0.88,
"learning_rate": 7.000000000000001e-06,
"loss": 0.1785,
"step": 2800
},
{
"epoch": 0.91,
"learning_rate": 7.25e-06,
"loss": 0.1725,
"step": 2900
},
{
"epoch": 0.94,
"learning_rate": 7.5e-06,
"loss": 0.1713,
"step": 3000
},
{
"epoch": 0.97,
"learning_rate": 7.75e-06,
"loss": 0.1684,
"step": 3100
},
{
"epoch": 1.0,
"learning_rate": 8.000000000000001e-06,
"loss": 0.1694,
"step": 3200
},
{
"epoch": 1.03,
"learning_rate": 8.25e-06,
"loss": 0.1642,
"step": 3300
},
{
"epoch": 1.07,
"learning_rate": 8.500000000000002e-06,
"loss": 0.1608,
"step": 3400
},
{
"epoch": 1.1,
"learning_rate": 8.75e-06,
"loss": 0.1578,
"step": 3500
},
{
"epoch": 1.13,
"learning_rate": 9e-06,
"loss": 0.1567,
"step": 3600
},
{
"epoch": 1.16,
"learning_rate": 9.25e-06,
"loss": 0.1552,
"step": 3700
},
{
"epoch": 1.19,
"learning_rate": 9.5e-06,
"loss": 0.1567,
"step": 3800
},
{
"epoch": 1.22,
"learning_rate": 9.750000000000002e-06,
"loss": 0.1557,
"step": 3900
},
{
"epoch": 1.25,
"learning_rate": 1e-05,
"loss": 0.1539,
"step": 4000
},
{
"epoch": 1.28,
"learning_rate": 1.025e-05,
"loss": 0.1541,
"step": 4100
},
{
"epoch": 1.32,
"learning_rate": 1.05e-05,
"loss": 0.1518,
"step": 4200
},
{
"epoch": 1.35,
"learning_rate": 1.075e-05,
"loss": 0.1513,
"step": 4300
},
{
"epoch": 1.38,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.1503,
"step": 4400
},
{
"epoch": 1.41,
"learning_rate": 1.125e-05,
"loss": 0.1506,
"step": 4500
},
{
"epoch": 1.44,
"learning_rate": 1.1500000000000002e-05,
"loss": 0.1519,
"step": 4600
},
{
"epoch": 1.47,
"learning_rate": 1.175e-05,
"loss": 0.152,
"step": 4700
},
{
"epoch": 1.5,
"learning_rate": 1.2e-05,
"loss": 0.1488,
"step": 4800
},
{
"epoch": 1.54,
"learning_rate": 1.225e-05,
"loss": 0.1471,
"step": 4900
},
{
"epoch": 1.57,
"learning_rate": 1.25e-05,
"loss": 0.1476,
"step": 5000
},
{
"epoch": 1.57,
"eval_loss": 0.2646298408508301,
"eval_runtime": 97.2387,
"eval_samples_per_second": 91.949,
"eval_steps_per_second": 1.44,
"step": 5000
},
{
"epoch": 1.6,
"learning_rate": 1.2478755948334468e-05,
"loss": 0.1465,
"step": 5100
},
{
"epoch": 1.63,
"learning_rate": 1.2457511896668933e-05,
"loss": 0.1463,
"step": 5200
},
{
"epoch": 1.66,
"learning_rate": 1.24362678450034e-05,
"loss": 0.1429,
"step": 5300
},
{
"epoch": 1.69,
"learning_rate": 1.2415023793337865e-05,
"loss": 0.1439,
"step": 5400
},
{
"epoch": 1.72,
"learning_rate": 1.2393779741672332e-05,
"loss": 0.1429,
"step": 5500
},
{
"epoch": 1.75,
"learning_rate": 1.23725356900068e-05,
"loss": 0.14,
"step": 5600
},
{
"epoch": 1.79,
"learning_rate": 1.2351291638341265e-05,
"loss": 0.1395,
"step": 5700
},
{
"epoch": 1.82,
"learning_rate": 1.2330047586675732e-05,
"loss": 0.139,
"step": 5800
},
{
"epoch": 1.85,
"learning_rate": 1.2308803535010199e-05,
"loss": 0.137,
"step": 5900
},
{
"epoch": 1.88,
"learning_rate": 1.2287559483344664e-05,
"loss": 0.1336,
"step": 6000
},
{
"epoch": 1.91,
"learning_rate": 1.2266315431679131e-05,
"loss": 0.1368,
"step": 6100
},
{
"epoch": 1.94,
"learning_rate": 1.2245071380013596e-05,
"loss": 0.1348,
"step": 6200
},
{
"epoch": 1.97,
"learning_rate": 1.2223827328348063e-05,
"loss": 0.1337,
"step": 6300
},
{
"epoch": 2.01,
"learning_rate": 1.2202583276682529e-05,
"loss": 0.1338,
"step": 6400
},
{
"epoch": 2.04,
"learning_rate": 1.2181339225016996e-05,
"loss": 0.1259,
"step": 6500
},
{
"epoch": 2.07,
"learning_rate": 1.2160095173351463e-05,
"loss": 0.124,
"step": 6600
},
{
"epoch": 2.1,
"learning_rate": 1.213885112168593e-05,
"loss": 0.1244,
"step": 6700
},
{
"epoch": 2.13,
"learning_rate": 1.2117607070020395e-05,
"loss": 0.123,
"step": 6800
},
{
"epoch": 2.16,
"learning_rate": 1.2096363018354862e-05,
"loss": 0.1206,
"step": 6900
},
{
"epoch": 2.19,
"learning_rate": 1.2075118966689327e-05,
"loss": 0.1229,
"step": 7000
},
{
"epoch": 2.22,
"learning_rate": 1.2053874915023794e-05,
"loss": 0.1204,
"step": 7100
},
{
"epoch": 2.26,
"learning_rate": 1.203263086335826e-05,
"loss": 0.1223,
"step": 7200
},
{
"epoch": 2.29,
"learning_rate": 1.2011386811692727e-05,
"loss": 0.12,
"step": 7300
},
{
"epoch": 2.32,
"learning_rate": 1.1990142760027192e-05,
"loss": 0.1193,
"step": 7400
},
{
"epoch": 2.35,
"learning_rate": 1.196889870836166e-05,
"loss": 0.117,
"step": 7500
},
{
"epoch": 2.38,
"learning_rate": 1.1947654656696125e-05,
"loss": 0.1184,
"step": 7600
},
{
"epoch": 2.41,
"learning_rate": 1.1926410605030593e-05,
"loss": 0.1187,
"step": 7700
},
{
"epoch": 2.44,
"learning_rate": 1.1905166553365059e-05,
"loss": 0.1171,
"step": 7800
},
{
"epoch": 2.48,
"learning_rate": 1.1883922501699526e-05,
"loss": 0.1156,
"step": 7900
},
{
"epoch": 2.51,
"learning_rate": 1.1862678450033991e-05,
"loss": 0.1171,
"step": 8000
},
{
"epoch": 2.54,
"learning_rate": 1.1841434398368458e-05,
"loss": 0.1176,
"step": 8100
},
{
"epoch": 2.57,
"learning_rate": 1.1820190346702923e-05,
"loss": 0.1149,
"step": 8200
},
{
"epoch": 2.6,
"learning_rate": 1.179894629503739e-05,
"loss": 0.1129,
"step": 8300
},
{
"epoch": 2.63,
"learning_rate": 1.1777702243371856e-05,
"loss": 0.1162,
"step": 8400
},
{
"epoch": 2.66,
"learning_rate": 1.1756458191706323e-05,
"loss": 0.1127,
"step": 8500
},
{
"epoch": 2.69,
"learning_rate": 1.1735214140040788e-05,
"loss": 0.1146,
"step": 8600
},
{
"epoch": 2.73,
"learning_rate": 1.1713970088375255e-05,
"loss": 0.1112,
"step": 8700
},
{
"epoch": 2.76,
"learning_rate": 1.1692726036709722e-05,
"loss": 0.1123,
"step": 8800
},
{
"epoch": 2.79,
"learning_rate": 1.1671481985044189e-05,
"loss": 0.1127,
"step": 8900
},
{
"epoch": 2.82,
"learning_rate": 1.1650237933378654e-05,
"loss": 0.1132,
"step": 9000
},
{
"epoch": 2.85,
"learning_rate": 1.1628993881713121e-05,
"loss": 0.1118,
"step": 9100
},
{
"epoch": 2.88,
"learning_rate": 1.1607749830047587e-05,
"loss": 0.1121,
"step": 9200
},
{
"epoch": 2.91,
"learning_rate": 1.1586505778382054e-05,
"loss": 0.1103,
"step": 9300
},
{
"epoch": 2.95,
"learning_rate": 1.1565261726716519e-05,
"loss": 0.1098,
"step": 9400
},
{
"epoch": 2.98,
"learning_rate": 1.1544017675050986e-05,
"loss": 0.1071,
"step": 9500
},
{
"epoch": 3.01,
"learning_rate": 1.1522773623385451e-05,
"loss": 0.1065,
"step": 9600
},
{
"epoch": 3.04,
"learning_rate": 1.1501529571719918e-05,
"loss": 0.1019,
"step": 9700
},
{
"epoch": 3.07,
"learning_rate": 1.1480285520054385e-05,
"loss": 0.1017,
"step": 9800
},
{
"epoch": 3.1,
"learning_rate": 1.1459041468388852e-05,
"loss": 0.1023,
"step": 9900
},
{
"epoch": 3.13,
"learning_rate": 1.1437797416723318e-05,
"loss": 0.0995,
"step": 10000
},
{
"epoch": 3.13,
"eval_loss": 0.21612465381622314,
"eval_runtime": 83.5806,
"eval_samples_per_second": 106.975,
"eval_steps_per_second": 1.675,
"step": 10000
},
{
"epoch": 3.16,
"learning_rate": 1.1416553365057785e-05,
"loss": 0.1016,
"step": 10100
},
{
"epoch": 3.2,
"learning_rate": 1.139530931339225e-05,
"loss": 0.0998,
"step": 10200
},
{
"epoch": 3.23,
"learning_rate": 1.1374065261726717e-05,
"loss": 0.1009,
"step": 10300
},
{
"epoch": 3.26,
"learning_rate": 1.1352821210061182e-05,
"loss": 0.0986,
"step": 10400
},
{
"epoch": 3.29,
"learning_rate": 1.133157715839565e-05,
"loss": 0.1003,
"step": 10500
},
{
"epoch": 3.32,
"learning_rate": 1.1310333106730117e-05,
"loss": 0.1005,
"step": 10600
},
{
"epoch": 3.35,
"learning_rate": 1.1289089055064582e-05,
"loss": 0.1012,
"step": 10700
},
{
"epoch": 3.38,
"learning_rate": 1.1267845003399049e-05,
"loss": 0.1005,
"step": 10800
},
{
"epoch": 3.42,
"learning_rate": 1.1246600951733516e-05,
"loss": 0.1,
"step": 10900
},
{
"epoch": 3.45,
"learning_rate": 1.1225356900067981e-05,
"loss": 0.0994,
"step": 11000
},
{
"epoch": 3.48,
"learning_rate": 1.1204112848402448e-05,
"loss": 0.0972,
"step": 11100
},
{
"epoch": 3.51,
"learning_rate": 1.1182868796736915e-05,
"loss": 0.0981,
"step": 11200
},
{
"epoch": 3.54,
"learning_rate": 1.116162474507138e-05,
"loss": 0.0956,
"step": 11300
},
{
"epoch": 3.57,
"learning_rate": 1.1140380693405848e-05,
"loss": 0.0955,
"step": 11400
},
{
"epoch": 3.6,
"learning_rate": 1.1119136641740313e-05,
"loss": 0.0955,
"step": 11500
},
{
"epoch": 3.63,
"learning_rate": 1.109789259007478e-05,
"loss": 0.0978,
"step": 11600
},
{
"epoch": 3.67,
"learning_rate": 1.1076648538409245e-05,
"loss": 0.0959,
"step": 11700
},
{
"epoch": 3.7,
"learning_rate": 1.1055404486743712e-05,
"loss": 0.0968,
"step": 11800
},
{
"epoch": 3.73,
"learning_rate": 1.1034160435078178e-05,
"loss": 0.096,
"step": 11900
},
{
"epoch": 3.76,
"learning_rate": 1.1012916383412646e-05,
"loss": 0.0959,
"step": 12000
},
{
"epoch": 3.79,
"learning_rate": 1.0991672331747112e-05,
"loss": 0.0959,
"step": 12100
},
{
"epoch": 3.82,
"learning_rate": 1.0970428280081579e-05,
"loss": 0.0959,
"step": 12200
},
{
"epoch": 3.85,
"learning_rate": 1.0949184228416044e-05,
"loss": 0.0957,
"step": 12300
},
{
"epoch": 3.89,
"learning_rate": 1.0927940176750511e-05,
"loss": 0.0948,
"step": 12400
},
{
"epoch": 3.92,
"learning_rate": 1.0906696125084976e-05,
"loss": 0.0959,
"step": 12500
},
{
"epoch": 3.95,
"learning_rate": 1.0885452073419443e-05,
"loss": 0.0942,
"step": 12600
},
{
"epoch": 3.98,
"learning_rate": 1.0864208021753909e-05,
"loss": 0.0941,
"step": 12700
},
{
"epoch": 4.01,
"learning_rate": 1.0842963970088376e-05,
"loss": 0.0905,
"step": 12800
},
{
"epoch": 4.04,
"learning_rate": 1.0821719918422841e-05,
"loss": 0.0887,
"step": 12900
},
{
"epoch": 4.07,
"learning_rate": 1.080047586675731e-05,
"loss": 0.0873,
"step": 13000
},
{
"epoch": 4.1,
"learning_rate": 1.0779231815091775e-05,
"loss": 0.0887,
"step": 13100
},
{
"epoch": 4.14,
"learning_rate": 1.0757987763426242e-05,
"loss": 0.0879,
"step": 13200
},
{
"epoch": 4.17,
"learning_rate": 1.0736743711760707e-05,
"loss": 0.0889,
"step": 13300
},
{
"epoch": 4.2,
"learning_rate": 1.0715499660095175e-05,
"loss": 0.0869,
"step": 13400
},
{
"epoch": 4.23,
"learning_rate": 1.069425560842964e-05,
"loss": 0.0865,
"step": 13500
},
{
"epoch": 4.26,
"learning_rate": 1.0673011556764107e-05,
"loss": 0.0863,
"step": 13600
},
{
"epoch": 4.29,
"learning_rate": 1.0651767505098572e-05,
"loss": 0.0873,
"step": 13700
},
{
"epoch": 4.32,
"learning_rate": 1.063052345343304e-05,
"loss": 0.0856,
"step": 13800
},
{
"epoch": 4.36,
"learning_rate": 1.0609279401767505e-05,
"loss": 0.0872,
"step": 13900
},
{
"epoch": 4.39,
"learning_rate": 1.0588035350101972e-05,
"loss": 0.0865,
"step": 14000
},
{
"epoch": 4.42,
"learning_rate": 1.0566791298436439e-05,
"loss": 0.0868,
"step": 14100
},
{
"epoch": 4.45,
"learning_rate": 1.0545547246770906e-05,
"loss": 0.0858,
"step": 14200
},
{
"epoch": 4.48,
"learning_rate": 1.0524303195105371e-05,
"loss": 0.0863,
"step": 14300
},
{
"epoch": 4.51,
"learning_rate": 1.0503059143439838e-05,
"loss": 0.0861,
"step": 14400
},
{
"epoch": 4.54,
"learning_rate": 1.0481815091774303e-05,
"loss": 0.0855,
"step": 14500
},
{
"epoch": 4.57,
"learning_rate": 1.046057104010877e-05,
"loss": 0.0874,
"step": 14600
},
{
"epoch": 4.61,
"learning_rate": 1.0439326988443236e-05,
"loss": 0.0858,
"step": 14700
},
{
"epoch": 4.64,
"learning_rate": 1.0418082936777703e-05,
"loss": 0.0848,
"step": 14800
},
{
"epoch": 4.67,
"learning_rate": 1.0396838885112168e-05,
"loss": 0.0846,
"step": 14900
},
{
"epoch": 4.7,
"learning_rate": 1.0375594833446635e-05,
"loss": 0.0835,
"step": 15000
},
{
"epoch": 4.7,
"eval_loss": 0.1958872377872467,
"eval_runtime": 88.1156,
"eval_samples_per_second": 101.469,
"eval_steps_per_second": 1.589,
"step": 15000
},
{
"epoch": 4.73,
"learning_rate": 1.03543507817811e-05,
"loss": 0.0854,
"step": 15100
},
{
"epoch": 4.76,
"learning_rate": 1.0333106730115569e-05,
"loss": 0.0835,
"step": 15200
},
{
"epoch": 4.79,
"learning_rate": 1.0311862678450034e-05,
"loss": 0.0849,
"step": 15300
},
{
"epoch": 4.83,
"learning_rate": 1.0290618626784501e-05,
"loss": 0.0825,
"step": 15400
},
{
"epoch": 4.86,
"learning_rate": 1.0269374575118967e-05,
"loss": 0.0859,
"step": 15500
},
{
"epoch": 4.89,
"learning_rate": 1.0248130523453434e-05,
"loss": 0.0837,
"step": 15600
},
{
"epoch": 4.92,
"learning_rate": 1.0226886471787899e-05,
"loss": 0.0843,
"step": 15700
},
{
"epoch": 4.95,
"learning_rate": 1.0205642420122366e-05,
"loss": 0.0826,
"step": 15800
},
{
"epoch": 4.98,
"learning_rate": 1.0184398368456833e-05,
"loss": 0.0838,
"step": 15900
},
{
"epoch": 5.01,
"learning_rate": 1.0163154316791298e-05,
"loss": 0.081,
"step": 16000
},
{
"epoch": 5.04,
"learning_rate": 1.0141910265125765e-05,
"loss": 0.0767,
"step": 16100
},
{
"epoch": 5.08,
"learning_rate": 1.0120666213460232e-05,
"loss": 0.0778,
"step": 16200
},
{
"epoch": 5.11,
"learning_rate": 1.0099422161794698e-05,
"loss": 0.0767,
"step": 16300
},
{
"epoch": 5.14,
"learning_rate": 1.0078178110129165e-05,
"loss": 0.0775,
"step": 16400
},
{
"epoch": 5.17,
"learning_rate": 1.005693405846363e-05,
"loss": 0.0774,
"step": 16500
},
{
"epoch": 5.2,
"learning_rate": 1.0035690006798097e-05,
"loss": 0.0775,
"step": 16600
},
{
"epoch": 5.23,
"learning_rate": 1.0014445955132564e-05,
"loss": 0.0778,
"step": 16700
},
{
"epoch": 5.26,
"learning_rate": 9.99320190346703e-06,
"loss": 0.078,
"step": 16800
},
{
"epoch": 5.3,
"learning_rate": 9.971957851801497e-06,
"loss": 0.0777,
"step": 16900
},
{
"epoch": 5.33,
"learning_rate": 9.950713800135962e-06,
"loss": 0.0793,
"step": 17000
},
{
"epoch": 5.36,
"learning_rate": 9.929469748470429e-06,
"loss": 0.0765,
"step": 17100
},
{
"epoch": 5.39,
"learning_rate": 9.908225696804894e-06,
"loss": 0.0771,
"step": 17200
},
{
"epoch": 5.42,
"learning_rate": 9.886981645139363e-06,
"loss": 0.0768,
"step": 17300
},
{
"epoch": 5.45,
"learning_rate": 9.865737593473828e-06,
"loss": 0.0761,
"step": 17400
},
{
"epoch": 5.48,
"learning_rate": 9.844493541808295e-06,
"loss": 0.0778,
"step": 17500
},
{
"epoch": 5.51,
"learning_rate": 9.82324949014276e-06,
"loss": 0.0787,
"step": 17600
},
{
"epoch": 5.55,
"learning_rate": 9.802005438477228e-06,
"loss": 0.0752,
"step": 17700
},
{
"epoch": 5.58,
"learning_rate": 9.780761386811693e-06,
"loss": 0.0756,
"step": 17800
},
{
"epoch": 5.61,
"learning_rate": 9.75951733514616e-06,
"loss": 0.0761,
"step": 17900
},
{
"epoch": 5.64,
"learning_rate": 9.738273283480625e-06,
"loss": 0.0776,
"step": 18000
},
{
"epoch": 5.67,
"learning_rate": 9.717029231815092e-06,
"loss": 0.0746,
"step": 18100
},
{
"epoch": 5.7,
"learning_rate": 9.695785180149558e-06,
"loss": 0.0765,
"step": 18200
},
{
"epoch": 5.73,
"learning_rate": 9.674541128484025e-06,
"loss": 0.0762,
"step": 18300
},
{
"epoch": 5.77,
"learning_rate": 9.653297076818492e-06,
"loss": 0.0763,
"step": 18400
},
{
"epoch": 5.8,
"learning_rate": 9.632053025152959e-06,
"loss": 0.0754,
"step": 18500
},
{
"epoch": 5.83,
"learning_rate": 9.610808973487424e-06,
"loss": 0.0772,
"step": 18600
},
{
"epoch": 5.86,
"learning_rate": 9.589564921821891e-06,
"loss": 0.0739,
"step": 18700
},
{
"epoch": 5.89,
"learning_rate": 9.568320870156356e-06,
"loss": 0.0732,
"step": 18800
},
{
"epoch": 5.92,
"learning_rate": 9.547076818490823e-06,
"loss": 0.0756,
"step": 18900
},
{
"epoch": 5.95,
"learning_rate": 9.525832766825289e-06,
"loss": 0.0757,
"step": 19000
},
{
"epoch": 5.98,
"learning_rate": 9.504588715159756e-06,
"loss": 0.0747,
"step": 19100
},
{
"epoch": 6.02,
"learning_rate": 9.483344663494221e-06,
"loss": 0.0721,
"step": 19200
},
{
"epoch": 6.05,
"learning_rate": 9.462100611828688e-06,
"loss": 0.0695,
"step": 19300
},
{
"epoch": 6.08,
"learning_rate": 9.440856560163155e-06,
"loss": 0.0694,
"step": 19400
},
{
"epoch": 6.11,
"learning_rate": 9.419612508497622e-06,
"loss": 0.0696,
"step": 19500
},
{
"epoch": 6.14,
"learning_rate": 9.398368456832087e-06,
"loss": 0.0695,
"step": 19600
},
{
"epoch": 6.17,
"learning_rate": 9.377124405166555e-06,
"loss": 0.069,
"step": 19700
},
{
"epoch": 6.2,
"learning_rate": 9.35588035350102e-06,
"loss": 0.0708,
"step": 19800
},
{
"epoch": 6.24,
"learning_rate": 9.334636301835487e-06,
"loss": 0.07,
"step": 19900
},
{
"epoch": 6.27,
"learning_rate": 9.313392250169952e-06,
"loss": 0.0706,
"step": 20000
},
{
"epoch": 6.27,
"eval_loss": 0.18451732397079468,
"eval_runtime": 91.4249,
"eval_samples_per_second": 97.796,
"eval_steps_per_second": 1.531,
"step": 20000
},
{
"epoch": 6.3,
"learning_rate": 9.29214819850442e-06,
"loss": 0.0692,
"step": 20100
},
{
"epoch": 6.33,
"learning_rate": 9.270904146838885e-06,
"loss": 0.0681,
"step": 20200
},
{
"epoch": 6.36,
"learning_rate": 9.249660095173352e-06,
"loss": 0.0712,
"step": 20300
},
{
"epoch": 6.39,
"learning_rate": 9.228416043507817e-06,
"loss": 0.0683,
"step": 20400
},
{
"epoch": 6.42,
"learning_rate": 9.207171991842286e-06,
"loss": 0.0684,
"step": 20500
},
{
"epoch": 6.45,
"learning_rate": 9.185927940176751e-06,
"loss": 0.0688,
"step": 20600
},
{
"epoch": 6.49,
"learning_rate": 9.164683888511218e-06,
"loss": 0.0687,
"step": 20700
},
{
"epoch": 6.52,
"learning_rate": 9.143439836845683e-06,
"loss": 0.0695,
"step": 20800
},
{
"epoch": 6.55,
"learning_rate": 9.12219578518015e-06,
"loss": 0.0695,
"step": 20900
},
{
"epoch": 6.58,
"learning_rate": 9.100951733514616e-06,
"loss": 0.069,
"step": 21000
},
{
"epoch": 6.61,
"learning_rate": 9.079707681849083e-06,
"loss": 0.0684,
"step": 21100
},
{
"epoch": 6.64,
"learning_rate": 9.058463630183548e-06,
"loss": 0.0675,
"step": 21200
},
{
"epoch": 6.67,
"learning_rate": 9.037219578518015e-06,
"loss": 0.07,
"step": 21300
},
{
"epoch": 6.71,
"learning_rate": 9.015975526852482e-06,
"loss": 0.0687,
"step": 21400
},
{
"epoch": 6.74,
"learning_rate": 8.994731475186947e-06,
"loss": 0.0683,
"step": 21500
},
{
"epoch": 6.77,
"learning_rate": 8.973487423521414e-06,
"loss": 0.0686,
"step": 21600
},
{
"epoch": 6.8,
"learning_rate": 8.952243371855881e-06,
"loss": 0.0687,
"step": 21700
},
{
"epoch": 6.83,
"learning_rate": 8.930999320190347e-06,
"loss": 0.0682,
"step": 21800
},
{
"epoch": 6.86,
"learning_rate": 8.909755268524814e-06,
"loss": 0.0698,
"step": 21900
},
{
"epoch": 6.89,
"learning_rate": 8.888511216859279e-06,
"loss": 0.068,
"step": 22000
},
{
"epoch": 6.92,
"learning_rate": 8.867267165193746e-06,
"loss": 0.068,
"step": 22100
},
{
"epoch": 6.96,
"learning_rate": 8.846023113528213e-06,
"loss": 0.0665,
"step": 22200
},
{
"epoch": 6.99,
"learning_rate": 8.824779061862678e-06,
"loss": 0.0701,
"step": 22300
},
{
"epoch": 7.02,
"learning_rate": 8.803535010197145e-06,
"loss": 0.0649,
"step": 22400
},
{
"epoch": 7.05,
"learning_rate": 8.78229095853161e-06,
"loss": 0.0632,
"step": 22500
},
{
"epoch": 7.08,
"learning_rate": 8.761046906866078e-06,
"loss": 0.0641,
"step": 22600
},
{
"epoch": 7.11,
"learning_rate": 8.739802855200545e-06,
"loss": 0.0611,
"step": 22700
},
{
"epoch": 7.14,
"learning_rate": 8.718558803535012e-06,
"loss": 0.0631,
"step": 22800
},
{
"epoch": 7.18,
"learning_rate": 8.697314751869477e-06,
"loss": 0.0644,
"step": 22900
},
{
"epoch": 7.21,
"learning_rate": 8.676070700203944e-06,
"loss": 0.0607,
"step": 23000
},
{
"epoch": 7.24,
"learning_rate": 8.65482664853841e-06,
"loss": 0.0627,
"step": 23100
},
{
"epoch": 7.27,
"learning_rate": 8.633582596872877e-06,
"loss": 0.0635,
"step": 23200
},
{
"epoch": 7.3,
"learning_rate": 8.612338545207342e-06,
"loss": 0.0635,
"step": 23300
},
{
"epoch": 7.33,
"learning_rate": 8.591094493541809e-06,
"loss": 0.0618,
"step": 23400
},
{
"epoch": 7.36,
"learning_rate": 8.569850441876274e-06,
"loss": 0.0625,
"step": 23500
},
{
"epoch": 7.39,
"learning_rate": 8.548606390210741e-06,
"loss": 0.0646,
"step": 23600
},
{
"epoch": 7.43,
"learning_rate": 8.527362338545208e-06,
"loss": 0.0628,
"step": 23700
},
{
"epoch": 7.46,
"learning_rate": 8.506118286879675e-06,
"loss": 0.063,
"step": 23800
},
{
"epoch": 7.49,
"learning_rate": 8.48487423521414e-06,
"loss": 0.0625,
"step": 23900
},
{
"epoch": 7.52,
"learning_rate": 8.463630183548608e-06,
"loss": 0.0632,
"step": 24000
},
{
"epoch": 7.55,
"learning_rate": 8.442386131883073e-06,
"loss": 0.0634,
"step": 24100
},
{
"epoch": 7.58,
"learning_rate": 8.42114208021754e-06,
"loss": 0.0622,
"step": 24200
},
{
"epoch": 7.61,
"learning_rate": 8.399898028552005e-06,
"loss": 0.0625,
"step": 24300
},
{
"epoch": 7.65,
"learning_rate": 8.378653976886472e-06,
"loss": 0.0632,
"step": 24400
},
{
"epoch": 7.68,
"learning_rate": 8.357409925220938e-06,
"loss": 0.0627,
"step": 24500
},
{
"epoch": 7.71,
"learning_rate": 8.336165873555405e-06,
"loss": 0.0631,
"step": 24600
},
{
"epoch": 7.74,
"learning_rate": 8.31492182188987e-06,
"loss": 0.0633,
"step": 24700
},
{
"epoch": 7.77,
"learning_rate": 8.293677770224339e-06,
"loss": 0.0638,
"step": 24800
},
{
"epoch": 7.8,
"learning_rate": 8.272433718558804e-06,
"loss": 0.0626,
"step": 24900
},
{
"epoch": 7.83,
"learning_rate": 8.251189666893271e-06,
"loss": 0.063,
"step": 25000
},
{
"epoch": 7.83,
"eval_loss": 0.17496153712272644,
"eval_runtime": 98.6326,
"eval_samples_per_second": 90.65,
"eval_steps_per_second": 1.419,
"step": 25000
},
{
"epoch": 7.86,
"learning_rate": 8.229945615227736e-06,
"loss": 0.0634,
"step": 25100
},
{
"epoch": 7.9,
"learning_rate": 8.208701563562203e-06,
"loss": 0.063,
"step": 25200
},
{
"epoch": 7.93,
"learning_rate": 8.187457511896669e-06,
"loss": 0.0634,
"step": 25300
},
{
"epoch": 7.96,
"learning_rate": 8.166213460231136e-06,
"loss": 0.0636,
"step": 25400
},
{
"epoch": 7.99,
"learning_rate": 8.144969408565601e-06,
"loss": 0.0607,
"step": 25500
},
{
"epoch": 8.02,
"learning_rate": 8.123725356900068e-06,
"loss": 0.0576,
"step": 25600
},
{
"epoch": 8.05,
"learning_rate": 8.102481305234533e-06,
"loss": 0.0579,
"step": 25700
},
{
"epoch": 8.08,
"learning_rate": 8.081237253569002e-06,
"loss": 0.0581,
"step": 25800
},
{
"epoch": 8.12,
"learning_rate": 8.059993201903467e-06,
"loss": 0.0584,
"step": 25900
},
{
"epoch": 8.15,
"learning_rate": 8.038749150237935e-06,
"loss": 0.0578,
"step": 26000
},
{
"epoch": 8.18,
"learning_rate": 8.0175050985724e-06,
"loss": 0.0564,
"step": 26100
},
{
"epoch": 8.21,
"learning_rate": 7.996261046906867e-06,
"loss": 0.0583,
"step": 26200
},
{
"epoch": 8.24,
"learning_rate": 7.975016995241332e-06,
"loss": 0.0568,
"step": 26300
},
{
"epoch": 8.27,
"learning_rate": 7.9537729435758e-06,
"loss": 0.0584,
"step": 26400
},
{
"epoch": 8.3,
"learning_rate": 7.932528891910265e-06,
"loss": 0.0585,
"step": 26500
},
{
"epoch": 8.33,
"learning_rate": 7.911284840244732e-06,
"loss": 0.0589,
"step": 26600
},
{
"epoch": 8.37,
"learning_rate": 7.890040788579197e-06,
"loss": 0.0569,
"step": 26700
},
{
"epoch": 8.4,
"learning_rate": 7.868796736913664e-06,
"loss": 0.0575,
"step": 26800
},
{
"epoch": 8.43,
"learning_rate": 7.847552685248131e-06,
"loss": 0.057,
"step": 26900
},
{
"epoch": 8.46,
"learning_rate": 7.826308633582598e-06,
"loss": 0.0591,
"step": 27000
},
{
"epoch": 8.49,
"learning_rate": 7.805064581917063e-06,
"loss": 0.0562,
"step": 27100
},
{
"epoch": 8.52,
"learning_rate": 7.78382053025153e-06,
"loss": 0.0573,
"step": 27200
},
{
"epoch": 8.55,
"learning_rate": 7.762576478585996e-06,
"loss": 0.0572,
"step": 27300
},
{
"epoch": 8.59,
"learning_rate": 7.741332426920463e-06,
"loss": 0.0583,
"step": 27400
},
{
"epoch": 8.62,
"learning_rate": 7.72008837525493e-06,
"loss": 0.0571,
"step": 27500
},
{
"epoch": 8.65,
"learning_rate": 7.698844323589395e-06,
"loss": 0.0565,
"step": 27600
},
{
"epoch": 8.68,
"learning_rate": 7.677600271923862e-06,
"loss": 0.0571,
"step": 27700
},
{
"epoch": 8.71,
"learning_rate": 7.656356220258327e-06,
"loss": 0.0561,
"step": 27800
},
{
"epoch": 8.74,
"learning_rate": 7.635112168592794e-06,
"loss": 0.0565,
"step": 27900
},
{
"epoch": 8.77,
"learning_rate": 7.613868116927261e-06,
"loss": 0.0572,
"step": 28000
},
{
"epoch": 8.8,
"learning_rate": 7.5926240652617276e-06,
"loss": 0.0568,
"step": 28100
},
{
"epoch": 8.84,
"learning_rate": 7.571380013596194e-06,
"loss": 0.0568,
"step": 28200
},
{
"epoch": 8.87,
"learning_rate": 7.55013596193066e-06,
"loss": 0.0569,
"step": 28300
},
{
"epoch": 8.9,
"learning_rate": 7.528891910265126e-06,
"loss": 0.0564,
"step": 28400
},
{
"epoch": 8.93,
"learning_rate": 7.507647858599592e-06,
"loss": 0.058,
"step": 28500
},
{
"epoch": 8.96,
"learning_rate": 7.4864038069340584e-06,
"loss": 0.0568,
"step": 28600
},
{
"epoch": 8.99,
"learning_rate": 7.465159755268525e-06,
"loss": 0.0577,
"step": 28700
},
{
"epoch": 9.02,
"learning_rate": 7.443915703602991e-06,
"loss": 0.0543,
"step": 28800
},
{
"epoch": 9.06,
"learning_rate": 7.422671651937457e-06,
"loss": 0.0535,
"step": 28900
},
{
"epoch": 9.09,
"learning_rate": 7.401427600271925e-06,
"loss": 0.0532,
"step": 29000
},
{
"epoch": 9.12,
"learning_rate": 7.380183548606391e-06,
"loss": 0.0527,
"step": 29100
},
{
"epoch": 9.15,
"learning_rate": 7.358939496940857e-06,
"loss": 0.0539,
"step": 29200
},
{
"epoch": 9.18,
"learning_rate": 7.337695445275323e-06,
"loss": 0.0519,
"step": 29300
},
{
"epoch": 9.21,
"learning_rate": 7.3164513936097895e-06,
"loss": 0.0513,
"step": 29400
},
{
"epoch": 9.24,
"learning_rate": 7.295207341944256e-06,
"loss": 0.0544,
"step": 29500
},
{
"epoch": 9.27,
"learning_rate": 7.273963290278722e-06,
"loss": 0.0541,
"step": 29600
},
{
"epoch": 9.31,
"learning_rate": 7.252719238613188e-06,
"loss": 0.0529,
"step": 29700
},
{
"epoch": 9.34,
"learning_rate": 7.231475186947654e-06,
"loss": 0.0536,
"step": 29800
},
{
"epoch": 9.37,
"learning_rate": 7.210231135282121e-06,
"loss": 0.0543,
"step": 29900
},
{
"epoch": 9.4,
"learning_rate": 7.1889870836165874e-06,
"loss": 0.0523,
"step": 30000
},
{
"epoch": 9.4,
"eval_loss": 0.16908282041549683,
"eval_runtime": 112.1691,
"eval_samples_per_second": 79.71,
"eval_steps_per_second": 1.248,
"step": 30000
},
{
"epoch": 9.43,
"learning_rate": 7.1677430319510544e-06,
"loss": 0.052,
"step": 30100
},
{
"epoch": 9.46,
"learning_rate": 7.146498980285521e-06,
"loss": 0.0524,
"step": 30200
},
{
"epoch": 9.49,
"learning_rate": 7.125254928619987e-06,
"loss": 0.0531,
"step": 30300
},
{
"epoch": 9.53,
"learning_rate": 7.104010876954453e-06,
"loss": 0.0521,
"step": 30400
},
{
"epoch": 9.56,
"learning_rate": 7.082766825288919e-06,
"loss": 0.0527,
"step": 30500
},
{
"epoch": 9.59,
"learning_rate": 7.061522773623386e-06,
"loss": 0.0519,
"step": 30600
},
{
"epoch": 9.62,
"learning_rate": 7.040278721957852e-06,
"loss": 0.0513,
"step": 30700
},
{
"epoch": 9.65,
"learning_rate": 7.0190346702923185e-06,
"loss": 0.0524,
"step": 30800
},
{
"epoch": 9.68,
"learning_rate": 6.997790618626785e-06,
"loss": 0.0543,
"step": 30900
},
{
"epoch": 9.71,
"learning_rate": 6.976546566961251e-06,
"loss": 0.0539,
"step": 31000
},
{
"epoch": 9.74,
"learning_rate": 6.955302515295717e-06,
"loss": 0.0512,
"step": 31100
},
{
"epoch": 9.78,
"learning_rate": 6.934058463630184e-06,
"loss": 0.0514,
"step": 31200
},
{
"epoch": 9.81,
"learning_rate": 6.912814411964651e-06,
"loss": 0.052,
"step": 31300
},
{
"epoch": 9.84,
"learning_rate": 6.891570360299117e-06,
"loss": 0.0524,
"step": 31400
},
{
"epoch": 9.87,
"learning_rate": 6.8703263086335834e-06,
"loss": 0.0525,
"step": 31500
},
{
"epoch": 9.9,
"learning_rate": 6.84908225696805e-06,
"loss": 0.0526,
"step": 31600
},
{
"epoch": 9.93,
"learning_rate": 6.827838205302516e-06,
"loss": 0.0523,
"step": 31700
},
{
"epoch": 9.96,
"learning_rate": 6.806594153636982e-06,
"loss": 0.0527,
"step": 31800
},
{
"epoch": 10.0,
"learning_rate": 6.785350101971448e-06,
"loss": 0.0525,
"step": 31900
},
{
"epoch": 10.03,
"learning_rate": 6.764106050305914e-06,
"loss": 0.0482,
"step": 32000
},
{
"epoch": 10.06,
"learning_rate": 6.7428619986403805e-06,
"loss": 0.0487,
"step": 32100
},
{
"epoch": 10.09,
"learning_rate": 6.721617946974848e-06,
"loss": 0.0497,
"step": 32200
},
{
"epoch": 10.12,
"learning_rate": 6.7003738953093145e-06,
"loss": 0.0474,
"step": 32300
},
{
"epoch": 10.15,
"learning_rate": 6.679129843643781e-06,
"loss": 0.0484,
"step": 32400
},
{
"epoch": 10.18,
"learning_rate": 6.657885791978247e-06,
"loss": 0.0491,
"step": 32500
},
{
"epoch": 10.21,
"learning_rate": 6.636641740312713e-06,
"loss": 0.0475,
"step": 32600
},
{
"epoch": 10.25,
"learning_rate": 6.615397688647179e-06,
"loss": 0.049,
"step": 32700
},
{
"epoch": 10.28,
"learning_rate": 6.594153636981645e-06,
"loss": 0.0477,
"step": 32800
},
{
"epoch": 10.31,
"learning_rate": 6.5729095853161116e-06,
"loss": 0.0474,
"step": 32900
},
{
"epoch": 10.34,
"learning_rate": 6.551665533650578e-06,
"loss": 0.0506,
"step": 33000
},
{
"epoch": 10.37,
"learning_rate": 6.530421481985044e-06,
"loss": 0.0479,
"step": 33100
},
{
"epoch": 10.4,
"learning_rate": 6.50917743031951e-06,
"loss": 0.0477,
"step": 33200
},
{
"epoch": 10.43,
"learning_rate": 6.487933378653978e-06,
"loss": 0.0489,
"step": 33300
},
{
"epoch": 10.47,
"learning_rate": 6.466689326988444e-06,
"loss": 0.0484,
"step": 33400
},
{
"epoch": 10.5,
"learning_rate": 6.44544527532291e-06,
"loss": 0.0498,
"step": 33500
},
{
"epoch": 10.53,
"learning_rate": 6.4242012236573765e-06,
"loss": 0.0485,
"step": 33600
},
{
"epoch": 10.56,
"learning_rate": 6.402957171991843e-06,
"loss": 0.0484,
"step": 33700
},
{
"epoch": 10.59,
"learning_rate": 6.381713120326309e-06,
"loss": 0.0468,
"step": 33800
},
{
"epoch": 10.62,
"learning_rate": 6.360469068660775e-06,
"loss": 0.0483,
"step": 33900
},
{
"epoch": 10.65,
"learning_rate": 6.339225016995241e-06,
"loss": 0.0493,
"step": 34000
},
{
"epoch": 10.68,
"learning_rate": 6.317980965329707e-06,
"loss": 0.048,
"step": 34100
},
{
"epoch": 10.72,
"learning_rate": 6.2967369136641735e-06,
"loss": 0.0505,
"step": 34200
},
{
"epoch": 10.75,
"learning_rate": 6.27549286199864e-06,
"loss": 0.0487,
"step": 34300
},
{
"epoch": 10.78,
"learning_rate": 6.2542488103331076e-06,
"loss": 0.0482,
"step": 34400
},
{
"epoch": 10.81,
"learning_rate": 6.233004758667573e-06,
"loss": 0.0469,
"step": 34500
},
{
"epoch": 10.84,
"learning_rate": 6.21176070700204e-06,
"loss": 0.0493,
"step": 34600
},
{
"epoch": 10.87,
"learning_rate": 6.190516655336506e-06,
"loss": 0.0488,
"step": 34700
},
{
"epoch": 10.9,
"learning_rate": 6.169272603670972e-06,
"loss": 0.0484,
"step": 34800
},
{
"epoch": 10.94,
"learning_rate": 6.1480285520054384e-06,
"loss": 0.0478,
"step": 34900
},
{
"epoch": 10.97,
"learning_rate": 6.126784500339905e-06,
"loss": 0.0497,
"step": 35000
},
{
"epoch": 10.97,
"eval_loss": 0.16288596391677856,
"eval_runtime": 95.7573,
"eval_samples_per_second": 93.372,
"eval_steps_per_second": 1.462,
"step": 35000
},
{
"epoch": 11.0,
"learning_rate": 6.105540448674371e-06,
"loss": 0.0481,
"step": 35100
},
{
"epoch": 11.03,
"learning_rate": 6.084296397008838e-06,
"loss": 0.0456,
"step": 35200
},
{
"epoch": 11.06,
"learning_rate": 6.063052345343304e-06,
"loss": 0.044,
"step": 35300
},
{
"epoch": 11.09,
"learning_rate": 6.04180829367777e-06,
"loss": 0.0443,
"step": 35400
},
{
"epoch": 11.12,
"learning_rate": 6.020564242012236e-06,
"loss": 0.045,
"step": 35500
},
{
"epoch": 11.15,
"learning_rate": 5.9993201903467025e-06,
"loss": 0.0435,
"step": 35600
},
{
"epoch": 11.19,
"learning_rate": 5.9780761386811695e-06,
"loss": 0.0448,
"step": 35700
},
{
"epoch": 11.22,
"learning_rate": 5.956832087015636e-06,
"loss": 0.0443,
"step": 35800
},
{
"epoch": 11.25,
"learning_rate": 5.935588035350102e-06,
"loss": 0.0447,
"step": 35900
},
{
"epoch": 11.28,
"learning_rate": 5.914343983684569e-06,
"loss": 0.0443,
"step": 36000
},
{
"epoch": 11.31,
"learning_rate": 5.893099932019035e-06,
"loss": 0.0454,
"step": 36100
},
{
"epoch": 11.34,
"learning_rate": 5.871855880353501e-06,
"loss": 0.0439,
"step": 36200
},
{
"epoch": 11.37,
"learning_rate": 5.8506118286879674e-06,
"loss": 0.0444,
"step": 36300
},
{
"epoch": 11.41,
"learning_rate": 5.8293677770224345e-06,
"loss": 0.0453,
"step": 36400
},
{
"epoch": 11.44,
"learning_rate": 5.808123725356901e-06,
"loss": 0.044,
"step": 36500
},
{
"epoch": 11.47,
"learning_rate": 5.786879673691367e-06,
"loss": 0.0449,
"step": 36600
},
{
"epoch": 11.5,
"learning_rate": 5.765635622025833e-06,
"loss": 0.045,
"step": 36700
},
{
"epoch": 11.53,
"learning_rate": 5.7443915703603e-06,
"loss": 0.0446,
"step": 36800
},
{
"epoch": 11.56,
"learning_rate": 5.723147518694766e-06,
"loss": 0.0453,
"step": 36900
},
{
"epoch": 11.59,
"learning_rate": 5.701903467029232e-06,
"loss": 0.0439,
"step": 37000
},
{
"epoch": 11.62,
"learning_rate": 5.6806594153636985e-06,
"loss": 0.0465,
"step": 37100
},
{
"epoch": 11.66,
"learning_rate": 5.659415363698165e-06,
"loss": 0.0446,
"step": 37200
},
{
"epoch": 11.69,
"learning_rate": 5.638171312032632e-06,
"loss": 0.0437,
"step": 37300
},
{
"epoch": 11.72,
"learning_rate": 5.616927260367098e-06,
"loss": 0.0446,
"step": 37400
},
{
"epoch": 11.75,
"learning_rate": 5.595683208701564e-06,
"loss": 0.0442,
"step": 37500
},
{
"epoch": 11.78,
"learning_rate": 5.57443915703603e-06,
"loss": 0.044,
"step": 37600
},
{
"epoch": 11.81,
"learning_rate": 5.553195105370496e-06,
"loss": 0.0446,
"step": 37700
},
{
"epoch": 11.84,
"learning_rate": 5.5319510537049634e-06,
"loss": 0.0443,
"step": 37800
},
{
"epoch": 11.88,
"learning_rate": 5.51070700203943e-06,
"loss": 0.0447,
"step": 37900
},
{
"epoch": 11.91,
"learning_rate": 5.489462950373896e-06,
"loss": 0.0446,
"step": 38000
},
{
"epoch": 11.94,
"learning_rate": 5.468218898708362e-06,
"loss": 0.0442,
"step": 38100
},
{
"epoch": 11.97,
"learning_rate": 5.446974847042828e-06,
"loss": 0.0464,
"step": 38200
},
{
"epoch": 12.0,
"learning_rate": 5.425730795377294e-06,
"loss": 0.0439,
"step": 38300
},
{
"epoch": 12.03,
"learning_rate": 5.404486743711761e-06,
"loss": 0.0408,
"step": 38400
},
{
"epoch": 12.06,
"learning_rate": 5.3832426920462275e-06,
"loss": 0.0407,
"step": 38500
},
{
"epoch": 12.09,
"learning_rate": 5.361998640380694e-06,
"loss": 0.0411,
"step": 38600
},
{
"epoch": 12.13,
"learning_rate": 5.34075458871516e-06,
"loss": 0.0417,
"step": 38700
},
{
"epoch": 12.16,
"learning_rate": 5.319510537049626e-06,
"loss": 0.042,
"step": 38800
},
{
"epoch": 12.19,
"learning_rate": 5.298266485384093e-06,
"loss": 0.0405,
"step": 38900
},
{
"epoch": 12.22,
"learning_rate": 5.277022433718559e-06,
"loss": 0.0414,
"step": 39000
},
{
"epoch": 12.25,
"learning_rate": 5.255778382053025e-06,
"loss": 0.042,
"step": 39100
},
{
"epoch": 12.28,
"learning_rate": 5.2345343303874916e-06,
"loss": 0.0417,
"step": 39200
},
{
"epoch": 12.31,
"learning_rate": 5.213290278721958e-06,
"loss": 0.0426,
"step": 39300
},
{
"epoch": 12.35,
"learning_rate": 5.192046227056425e-06,
"loss": 0.0403,
"step": 39400
},
{
"epoch": 12.38,
"learning_rate": 5.170802175390891e-06,
"loss": 0.0397,
"step": 39500
},
{
"epoch": 12.41,
"learning_rate": 5.149558123725357e-06,
"loss": 0.0413,
"step": 39600
},
{
"epoch": 12.44,
"learning_rate": 5.128314072059823e-06,
"loss": 0.0415,
"step": 39700
},
{
"epoch": 12.47,
"learning_rate": 5.1070700203942895e-06,
"loss": 0.0403,
"step": 39800
},
{
"epoch": 12.5,
"learning_rate": 5.085825968728756e-06,
"loss": 0.0403,
"step": 39900
},
{
"epoch": 12.53,
"learning_rate": 5.064581917063223e-06,
"loss": 0.0407,
"step": 40000
},
{
"epoch": 12.53,
"eval_loss": 0.16060428321361542,
"eval_runtime": 83.676,
"eval_samples_per_second": 106.853,
"eval_steps_per_second": 1.673,
"step": 40000
},
{
"epoch": 12.56,
"learning_rate": 5.043337865397689e-06,
"loss": 0.0414,
"step": 40100
},
{
"epoch": 12.6,
"learning_rate": 5.022093813732155e-06,
"loss": 0.0414,
"step": 40200
},
{
"epoch": 12.63,
"learning_rate": 5.000849762066621e-06,
"loss": 0.0405,
"step": 40300
},
{
"epoch": 12.66,
"learning_rate": 4.979605710401087e-06,
"loss": 0.0407,
"step": 40400
},
{
"epoch": 12.69,
"learning_rate": 4.958361658735554e-06,
"loss": 0.0405,
"step": 40500
},
{
"epoch": 12.72,
"learning_rate": 4.9371176070700206e-06,
"loss": 0.0402,
"step": 40600
},
{
"epoch": 12.75,
"learning_rate": 4.915873555404487e-06,
"loss": 0.0416,
"step": 40700
},
{
"epoch": 12.78,
"learning_rate": 4.894629503738953e-06,
"loss": 0.0413,
"step": 40800
},
{
"epoch": 12.82,
"learning_rate": 4.873385452073419e-06,
"loss": 0.0411,
"step": 40900
},
{
"epoch": 12.85,
"learning_rate": 4.852141400407886e-06,
"loss": 0.0398,
"step": 41000
},
{
"epoch": 12.88,
"learning_rate": 4.830897348742352e-06,
"loss": 0.0406,
"step": 41100
},
{
"epoch": 12.91,
"learning_rate": 4.8096532970768185e-06,
"loss": 0.0409,
"step": 41200
},
{
"epoch": 12.94,
"learning_rate": 4.788409245411285e-06,
"loss": 0.0419,
"step": 41300
},
{
"epoch": 12.97,
"learning_rate": 4.767165193745751e-06,
"loss": 0.0402,
"step": 41400
},
{
"epoch": 13.0,
"learning_rate": 4.745921142080218e-06,
"loss": 0.0414,
"step": 41500
},
{
"epoch": 13.03,
"learning_rate": 4.724677090414684e-06,
"loss": 0.0384,
"step": 41600
},
{
"epoch": 13.07,
"learning_rate": 4.70343303874915e-06,
"loss": 0.0383,
"step": 41700
},
{
"epoch": 13.1,
"learning_rate": 4.682188987083617e-06,
"loss": 0.0375,
"step": 41800
},
{
"epoch": 13.13,
"learning_rate": 4.660944935418083e-06,
"loss": 0.0378,
"step": 41900
},
{
"epoch": 13.16,
"learning_rate": 4.6397008837525495e-06,
"loss": 0.0373,
"step": 42000
},
{
"epoch": 13.19,
"learning_rate": 4.618456832087016e-06,
"loss": 0.0367,
"step": 42100
},
{
"epoch": 13.22,
"learning_rate": 4.597212780421483e-06,
"loss": 0.0393,
"step": 42200
},
{
"epoch": 13.25,
"learning_rate": 4.575968728755949e-06,
"loss": 0.0381,
"step": 42300
},
{
"epoch": 13.29,
"learning_rate": 4.554724677090415e-06,
"loss": 0.0392,
"step": 42400
},
{
"epoch": 13.32,
"learning_rate": 4.533480625424881e-06,
"loss": 0.0371,
"step": 42500
},
{
"epoch": 13.35,
"learning_rate": 4.512236573759348e-06,
"loss": 0.0395,
"step": 42600
},
{
"epoch": 13.38,
"learning_rate": 4.4909925220938145e-06,
"loss": 0.0383,
"step": 42700
},
{
"epoch": 13.41,
"learning_rate": 4.469748470428281e-06,
"loss": 0.037,
"step": 42800
},
{
"epoch": 13.44,
"learning_rate": 4.448504418762747e-06,
"loss": 0.0371,
"step": 42900
},
{
"epoch": 13.47,
"learning_rate": 4.427260367097213e-06,
"loss": 0.0369,
"step": 43000
},
{
"epoch": 13.5,
"learning_rate": 4.406016315431679e-06,
"loss": 0.039,
"step": 43100
},
{
"epoch": 13.54,
"learning_rate": 4.384772263766146e-06,
"loss": 0.0365,
"step": 43200
},
{
"epoch": 13.57,
"learning_rate": 4.363528212100612e-06,
"loss": 0.0379,
"step": 43300
},
{
"epoch": 13.6,
"learning_rate": 4.3422841604350785e-06,
"loss": 0.0387,
"step": 43400
},
{
"epoch": 13.63,
"learning_rate": 4.321040108769545e-06,
"loss": 0.0383,
"step": 43500
},
{
"epoch": 13.66,
"learning_rate": 4.299796057104011e-06,
"loss": 0.0375,
"step": 43600
},
{
"epoch": 13.69,
"learning_rate": 4.278552005438478e-06,
"loss": 0.0376,
"step": 43700
},
{
"epoch": 13.72,
"learning_rate": 4.257307953772944e-06,
"loss": 0.0376,
"step": 43800
},
{
"epoch": 13.76,
"learning_rate": 4.23606390210741e-06,
"loss": 0.0377,
"step": 43900
},
{
"epoch": 13.79,
"learning_rate": 4.214819850441876e-06,
"loss": 0.0379,
"step": 44000
},
{
"epoch": 13.82,
"learning_rate": 4.193575798776343e-06,
"loss": 0.0377,
"step": 44100
},
{
"epoch": 13.85,
"learning_rate": 4.17233174711081e-06,
"loss": 0.037,
"step": 44200
},
{
"epoch": 13.88,
"learning_rate": 4.151087695445276e-06,
"loss": 0.0382,
"step": 44300
},
{
"epoch": 13.91,
"learning_rate": 4.129843643779742e-06,
"loss": 0.0396,
"step": 44400
},
{
"epoch": 13.94,
"learning_rate": 4.108599592114208e-06,
"loss": 0.0377,
"step": 44500
},
{
"epoch": 13.97,
"learning_rate": 4.087355540448674e-06,
"loss": 0.0381,
"step": 44600
},
{
"epoch": 14.01,
"learning_rate": 4.0661114887831405e-06,
"loss": 0.0375,
"step": 44700
},
{
"epoch": 14.04,
"learning_rate": 4.0448674371176075e-06,
"loss": 0.0348,
"step": 44800
},
{
"epoch": 14.07,
"learning_rate": 4.023623385452074e-06,
"loss": 0.0349,
"step": 44900
},
{
"epoch": 14.1,
"learning_rate": 4.00237933378654e-06,
"loss": 0.0345,
"step": 45000
},
{
"epoch": 14.1,
"eval_loss": 0.15830282866954803,
"eval_runtime": 116.727,
"eval_samples_per_second": 76.598,
"eval_steps_per_second": 1.199,
"step": 45000
},
{
"epoch": 14.13,
"learning_rate": 3.981135282121006e-06,
"loss": 0.0347,
"step": 45100
},
{
"epoch": 14.16,
"learning_rate": 3.959891230455472e-06,
"loss": 0.0354,
"step": 45200
},
{
"epoch": 14.19,
"learning_rate": 3.938647178789939e-06,
"loss": 0.0349,
"step": 45300
},
{
"epoch": 14.23,
"learning_rate": 3.917403127124405e-06,
"loss": 0.0346,
"step": 45400
},
{
"epoch": 14.26,
"learning_rate": 3.896159075458872e-06,
"loss": 0.0352,
"step": 45500
},
{
"epoch": 14.29,
"learning_rate": 3.874915023793338e-06,
"loss": 0.0349,
"step": 45600
},
{
"epoch": 14.32,
"learning_rate": 3.853670972127804e-06,
"loss": 0.0358,
"step": 45700
},
{
"epoch": 14.35,
"learning_rate": 3.832426920462271e-06,
"loss": 0.0351,
"step": 45800
},
{
"epoch": 14.38,
"learning_rate": 3.811182868796737e-06,
"loss": 0.0344,
"step": 45900
},
{
"epoch": 14.41,
"learning_rate": 3.7899388171312033e-06,
"loss": 0.0354,
"step": 46000
},
{
"epoch": 14.44,
"learning_rate": 3.76869476546567e-06,
"loss": 0.0355,
"step": 46100
},
{
"epoch": 14.48,
"learning_rate": 3.747450713800136e-06,
"loss": 0.0352,
"step": 46200
},
{
"epoch": 14.51,
"learning_rate": 3.7262066621346022e-06,
"loss": 0.035,
"step": 46300
},
{
"epoch": 14.54,
"learning_rate": 3.7049626104690693e-06,
"loss": 0.0349,
"step": 46400
},
{
"epoch": 14.57,
"learning_rate": 3.6837185588035354e-06,
"loss": 0.0349,
"step": 46500
},
{
"epoch": 14.6,
"learning_rate": 3.6624745071380016e-06,
"loss": 0.0333,
"step": 46600
},
{
"epoch": 14.63,
"learning_rate": 3.641230455472468e-06,
"loss": 0.0331,
"step": 46700
},
{
"epoch": 14.66,
"learning_rate": 3.619986403806934e-06,
"loss": 0.0351,
"step": 46800
},
{
"epoch": 14.7,
"learning_rate": 3.598742352141401e-06,
"loss": 0.0335,
"step": 46900
},
{
"epoch": 14.73,
"learning_rate": 3.577498300475867e-06,
"loss": 0.0349,
"step": 47000
},
{
"epoch": 14.76,
"learning_rate": 3.5562542488103333e-06,
"loss": 0.0352,
"step": 47100
},
{
"epoch": 14.79,
"learning_rate": 3.5350101971447995e-06,
"loss": 0.034,
"step": 47200
},
{
"epoch": 14.82,
"learning_rate": 3.5137661454792657e-06,
"loss": 0.0348,
"step": 47300
},
{
"epoch": 14.85,
"learning_rate": 3.4925220938137327e-06,
"loss": 0.0345,
"step": 47400
},
{
"epoch": 14.88,
"learning_rate": 3.471278042148199e-06,
"loss": 0.0345,
"step": 47500
},
{
"epoch": 14.91,
"learning_rate": 3.450033990482665e-06,
"loss": 0.0353,
"step": 47600
},
{
"epoch": 14.95,
"learning_rate": 3.4287899388171312e-06,
"loss": 0.0336,
"step": 47700
},
{
"epoch": 14.98,
"learning_rate": 3.4075458871515974e-06,
"loss": 0.035,
"step": 47800
},
{
"epoch": 15.01,
"learning_rate": 3.3863018354860636e-06,
"loss": 0.0348,
"step": 47900
},
{
"epoch": 15.04,
"learning_rate": 3.3650577838205306e-06,
"loss": 0.0323,
"step": 48000
},
{
"epoch": 15.07,
"learning_rate": 3.3438137321549968e-06,
"loss": 0.0319,
"step": 48100
},
{
"epoch": 15.1,
"learning_rate": 3.322569680489463e-06,
"loss": 0.0325,
"step": 48200
},
{
"epoch": 15.13,
"learning_rate": 3.301325628823929e-06,
"loss": 0.0324,
"step": 48300
},
{
"epoch": 15.17,
"learning_rate": 3.2800815771583953e-06,
"loss": 0.0337,
"step": 48400
},
{
"epoch": 15.2,
"learning_rate": 3.2588375254928623e-06,
"loss": 0.0322,
"step": 48500
},
{
"epoch": 15.23,
"learning_rate": 3.2375934738273285e-06,
"loss": 0.0321,
"step": 48600
},
{
"epoch": 15.26,
"learning_rate": 3.2163494221617947e-06,
"loss": 0.0327,
"step": 48700
},
{
"epoch": 15.29,
"learning_rate": 3.1951053704962613e-06,
"loss": 0.0322,
"step": 48800
},
{
"epoch": 15.32,
"learning_rate": 3.1738613188307274e-06,
"loss": 0.0334,
"step": 48900
},
{
"epoch": 15.35,
"learning_rate": 3.152617267165194e-06,
"loss": 0.0321,
"step": 49000
},
{
"epoch": 15.38,
"learning_rate": 3.1313732154996602e-06,
"loss": 0.0311,
"step": 49100
},
{
"epoch": 15.42,
"learning_rate": 3.110129163834127e-06,
"loss": 0.0322,
"step": 49200
},
{
"epoch": 15.45,
"learning_rate": 3.088885112168593e-06,
"loss": 0.0317,
"step": 49300
},
{
"epoch": 15.48,
"learning_rate": 3.0676410605030596e-06,
"loss": 0.0325,
"step": 49400
},
{
"epoch": 15.51,
"learning_rate": 3.0463970088375258e-06,
"loss": 0.0329,
"step": 49500
},
{
"epoch": 15.54,
"learning_rate": 3.025152957171992e-06,
"loss": 0.032,
"step": 49600
},
{
"epoch": 15.57,
"learning_rate": 3.0039089055064585e-06,
"loss": 0.031,
"step": 49700
},
{
"epoch": 15.6,
"learning_rate": 2.9826648538409247e-06,
"loss": 0.0312,
"step": 49800
},
{
"epoch": 15.64,
"learning_rate": 2.961420802175391e-06,
"loss": 0.0319,
"step": 49900
},
{
"epoch": 15.67,
"learning_rate": 2.9401767505098575e-06,
"loss": 0.0307,
"step": 50000
},
{
"epoch": 15.67,
"eval_loss": 0.15686601400375366,
"eval_runtime": 104.9673,
"eval_samples_per_second": 85.179,
"eval_steps_per_second": 1.334,
"step": 50000
},
{
"epoch": 15.7,
"learning_rate": 2.9189326988443237e-06,
"loss": 0.0317,
"step": 50100
},
{
"epoch": 15.73,
"learning_rate": 2.8976886471787902e-06,
"loss": 0.0312,
"step": 50200
},
{
"epoch": 15.76,
"learning_rate": 2.8764445955132564e-06,
"loss": 0.0305,
"step": 50300
},
{
"epoch": 15.79,
"learning_rate": 2.8552005438477226e-06,
"loss": 0.033,
"step": 50400
},
{
"epoch": 15.82,
"learning_rate": 2.833956492182189e-06,
"loss": 0.0318,
"step": 50500
},
{
"epoch": 15.85,
"learning_rate": 2.8127124405166554e-06,
"loss": 0.0312,
"step": 50600
},
{
"epoch": 15.89,
"learning_rate": 2.7914683888511215e-06,
"loss": 0.0318,
"step": 50700
},
{
"epoch": 15.92,
"learning_rate": 2.770224337185588e-06,
"loss": 0.0314,
"step": 50800
},
{
"epoch": 15.95,
"learning_rate": 2.7489802855200543e-06,
"loss": 0.0328,
"step": 50900
},
{
"epoch": 15.98,
"learning_rate": 2.727736233854521e-06,
"loss": 0.0325,
"step": 51000
},
{
"epoch": 16.01,
"learning_rate": 2.706492182188987e-06,
"loss": 0.0312,
"step": 51100
},
{
"epoch": 16.04,
"learning_rate": 2.6852481305234533e-06,
"loss": 0.0311,
"step": 51200
},
{
"epoch": 16.07,
"learning_rate": 2.66400407885792e-06,
"loss": 0.0298,
"step": 51300
},
{
"epoch": 16.11,
"learning_rate": 2.642760027192386e-06,
"loss": 0.0297,
"step": 51400
},
{
"epoch": 16.14,
"learning_rate": 2.6215159755268526e-06,
"loss": 0.03,
"step": 51500
},
{
"epoch": 16.17,
"learning_rate": 2.600271923861319e-06,
"loss": 0.0306,
"step": 51600
},
{
"epoch": 16.2,
"learning_rate": 2.5790278721957854e-06,
"loss": 0.0289,
"step": 51700
},
{
"epoch": 16.23,
"learning_rate": 2.5577838205302516e-06,
"loss": 0.0288,
"step": 51800
},
{
"epoch": 16.26,
"learning_rate": 2.536539768864718e-06,
"loss": 0.0303,
"step": 51900
},
{
"epoch": 16.29,
"learning_rate": 2.5152957171991844e-06,
"loss": 0.0289,
"step": 52000
},
{
"epoch": 16.32,
"learning_rate": 2.494051665533651e-06,
"loss": 0.0285,
"step": 52100
},
{
"epoch": 16.36,
"learning_rate": 2.472807613868117e-06,
"loss": 0.0303,
"step": 52200
},
{
"epoch": 16.39,
"learning_rate": 2.4515635622025833e-06,
"loss": 0.0296,
"step": 52300
},
{
"epoch": 16.42,
"learning_rate": 2.43031951053705e-06,
"loss": 0.0301,
"step": 52400
},
{
"epoch": 16.45,
"learning_rate": 2.409075458871516e-06,
"loss": 0.0285,
"step": 52500
},
{
"epoch": 16.48,
"learning_rate": 2.3878314072059827e-06,
"loss": 0.0289,
"step": 52600
},
{
"epoch": 16.51,
"learning_rate": 2.366587355540449e-06,
"loss": 0.0313,
"step": 52700
},
{
"epoch": 16.54,
"learning_rate": 2.345343303874915e-06,
"loss": 0.0307,
"step": 52800
},
{
"epoch": 16.58,
"learning_rate": 2.3240992522093816e-06,
"loss": 0.0293,
"step": 52900
},
{
"epoch": 16.61,
"learning_rate": 2.302855200543848e-06,
"loss": 0.0307,
"step": 53000
},
{
"epoch": 16.64,
"learning_rate": 2.281611148878314e-06,
"loss": 0.0296,
"step": 53100
},
{
"epoch": 16.67,
"learning_rate": 2.2603670972127806e-06,
"loss": 0.0295,
"step": 53200
},
{
"epoch": 16.7,
"learning_rate": 2.2391230455472467e-06,
"loss": 0.0288,
"step": 53300
},
{
"epoch": 16.73,
"learning_rate": 2.2178789938817133e-06,
"loss": 0.0299,
"step": 53400
},
{
"epoch": 16.76,
"learning_rate": 2.1966349422161795e-06,
"loss": 0.029,
"step": 53500
},
{
"epoch": 16.79,
"learning_rate": 2.1753908905506457e-06,
"loss": 0.0291,
"step": 53600
},
{
"epoch": 16.83,
"learning_rate": 2.1541468388851123e-06,
"loss": 0.0286,
"step": 53700
},
{
"epoch": 16.86,
"learning_rate": 2.1329027872195785e-06,
"loss": 0.0287,
"step": 53800
},
{
"epoch": 16.89,
"learning_rate": 2.1116587355540446e-06,
"loss": 0.0293,
"step": 53900
},
{
"epoch": 16.92,
"learning_rate": 2.0904146838885112e-06,
"loss": 0.0292,
"step": 54000
},
{
"epoch": 16.95,
"learning_rate": 2.0691706322229774e-06,
"loss": 0.0295,
"step": 54100
},
{
"epoch": 16.98,
"learning_rate": 2.047926580557444e-06,
"loss": 0.0299,
"step": 54200
},
{
"epoch": 17.01,
"learning_rate": 2.02668252889191e-06,
"loss": 0.0287,
"step": 54300
},
{
"epoch": 17.05,
"learning_rate": 2.0054384772263768e-06,
"loss": 0.0292,
"step": 54400
},
{
"epoch": 17.08,
"learning_rate": 1.984194425560843e-06,
"loss": 0.0272,
"step": 54500
},
{
"epoch": 17.11,
"learning_rate": 1.9629503738953096e-06,
"loss": 0.0271,
"step": 54600
},
{
"epoch": 17.14,
"learning_rate": 1.9417063222297757e-06,
"loss": 0.0275,
"step": 54700
},
{
"epoch": 17.17,
"learning_rate": 1.9204622705642423e-06,
"loss": 0.027,
"step": 54800
},
{
"epoch": 17.2,
"learning_rate": 1.8992182188987083e-06,
"loss": 0.0282,
"step": 54900
},
{
"epoch": 17.23,
"learning_rate": 1.8779741672331749e-06,
"loss": 0.0266,
"step": 55000
},
{
"epoch": 17.23,
"eval_loss": 0.15546093881130219,
"eval_runtime": 80.0972,
"eval_samples_per_second": 111.627,
"eval_steps_per_second": 1.748,
"step": 55000
},
{
"epoch": 17.26,
"learning_rate": 1.856730115567641e-06,
"loss": 0.028,
"step": 55100
},
{
"epoch": 17.3,
"learning_rate": 1.8354860639021074e-06,
"loss": 0.0274,
"step": 55200
},
{
"epoch": 17.33,
"learning_rate": 1.8142420122365738e-06,
"loss": 0.0279,
"step": 55300
},
{
"epoch": 17.36,
"learning_rate": 1.7929979605710402e-06,
"loss": 0.0288,
"step": 55400
},
{
"epoch": 17.39,
"learning_rate": 1.7717539089055068e-06,
"loss": 0.027,
"step": 55500
},
{
"epoch": 17.42,
"learning_rate": 1.750509857239973e-06,
"loss": 0.0267,
"step": 55600
},
{
"epoch": 17.45,
"learning_rate": 1.7292658055744392e-06,
"loss": 0.0288,
"step": 55700
},
{
"epoch": 17.48,
"learning_rate": 1.7080217539089058e-06,
"loss": 0.027,
"step": 55800
},
{
"epoch": 17.52,
"learning_rate": 1.686777702243372e-06,
"loss": 0.027,
"step": 55900
},
{
"epoch": 17.55,
"learning_rate": 1.6655336505778381e-06,
"loss": 0.0272,
"step": 56000
},
{
"epoch": 17.58,
"learning_rate": 1.6442895989123047e-06,
"loss": 0.0274,
"step": 56100
},
{
"epoch": 17.61,
"learning_rate": 1.6230455472467709e-06,
"loss": 0.0272,
"step": 56200
},
{
"epoch": 17.64,
"learning_rate": 1.6018014955812375e-06,
"loss": 0.0282,
"step": 56300
},
{
"epoch": 17.67,
"learning_rate": 1.5805574439157037e-06,
"loss": 0.0259,
"step": 56400
},
{
"epoch": 17.7,
"learning_rate": 1.55931339225017e-06,
"loss": 0.028,
"step": 56500
},
{
"epoch": 17.73,
"learning_rate": 1.5380693405846364e-06,
"loss": 0.0276,
"step": 56600
},
{
"epoch": 17.77,
"learning_rate": 1.5168252889191028e-06,
"loss": 0.0277,
"step": 56700
},
{
"epoch": 17.8,
"learning_rate": 1.4955812372535692e-06,
"loss": 0.0263,
"step": 56800
},
{
"epoch": 17.83,
"learning_rate": 1.4743371855880356e-06,
"loss": 0.0267,
"step": 56900
},
{
"epoch": 17.86,
"learning_rate": 1.4530931339225018e-06,
"loss": 0.0275,
"step": 57000
},
{
"epoch": 17.89,
"learning_rate": 1.4318490822569681e-06,
"loss": 0.027,
"step": 57100
},
{
"epoch": 17.92,
"learning_rate": 1.4106050305914345e-06,
"loss": 0.0268,
"step": 57200
},
{
"epoch": 17.95,
"learning_rate": 1.389360978925901e-06,
"loss": 0.0277,
"step": 57300
},
{
"epoch": 17.99,
"learning_rate": 1.368116927260367e-06,
"loss": 0.0268,
"step": 57400
},
{
"epoch": 18.02,
"learning_rate": 1.3468728755948335e-06,
"loss": 0.0268,
"step": 57500
},
{
"epoch": 18.05,
"learning_rate": 1.3256288239292999e-06,
"loss": 0.0253,
"step": 57600
},
{
"epoch": 18.08,
"learning_rate": 1.3043847722637663e-06,
"loss": 0.0253,
"step": 57700
},
{
"epoch": 18.11,
"learning_rate": 1.2831407205982324e-06,
"loss": 0.026,
"step": 57800
},
{
"epoch": 18.14,
"learning_rate": 1.2618966689326988e-06,
"loss": 0.0255,
"step": 57900
},
{
"epoch": 18.17,
"learning_rate": 1.2406526172671652e-06,
"loss": 0.0271,
"step": 58000
},
{
"epoch": 18.2,
"learning_rate": 1.2194085656016316e-06,
"loss": 0.0264,
"step": 58100
},
{
"epoch": 18.24,
"learning_rate": 1.198164513936098e-06,
"loss": 0.0266,
"step": 58200
},
{
"epoch": 18.27,
"learning_rate": 1.1769204622705644e-06,
"loss": 0.0269,
"step": 58300
},
{
"epoch": 18.3,
"learning_rate": 1.1556764106050307e-06,
"loss": 0.0251,
"step": 58400
},
{
"epoch": 18.33,
"learning_rate": 1.1344323589394971e-06,
"loss": 0.026,
"step": 58500
},
{
"epoch": 18.36,
"learning_rate": 1.1131883072739633e-06,
"loss": 0.0249,
"step": 58600
},
{
"epoch": 18.39,
"learning_rate": 1.0919442556084297e-06,
"loss": 0.0264,
"step": 58700
},
{
"epoch": 18.42,
"learning_rate": 1.070700203942896e-06,
"loss": 0.025,
"step": 58800
},
{
"epoch": 18.46,
"learning_rate": 1.0494561522773625e-06,
"loss": 0.0259,
"step": 58900
},
{
"epoch": 18.49,
"learning_rate": 1.0282121006118286e-06,
"loss": 0.0266,
"step": 59000
},
{
"epoch": 18.52,
"learning_rate": 1.006968048946295e-06,
"loss": 0.0251,
"step": 59100
},
{
"epoch": 18.55,
"learning_rate": 9.857239972807614e-07,
"loss": 0.0265,
"step": 59200
},
{
"epoch": 18.58,
"learning_rate": 9.644799456152278e-07,
"loss": 0.026,
"step": 59300
},
{
"epoch": 18.61,
"learning_rate": 9.432358939496941e-07,
"loss": 0.0254,
"step": 59400
},
{
"epoch": 18.64,
"learning_rate": 9.219918422841605e-07,
"loss": 0.0251,
"step": 59500
},
{
"epoch": 18.67,
"learning_rate": 9.007477906186269e-07,
"loss": 0.025,
"step": 59600
},
{
"epoch": 18.71,
"learning_rate": 8.795037389530932e-07,
"loss": 0.0247,
"step": 59700
},
{
"epoch": 18.74,
"learning_rate": 8.582596872875595e-07,
"loss": 0.0265,
"step": 59800
},
{
"epoch": 18.77,
"learning_rate": 8.370156356220259e-07,
"loss": 0.0259,
"step": 59900
},
{
"epoch": 18.8,
"learning_rate": 8.157715839564923e-07,
"loss": 0.0244,
"step": 60000
},
{
"epoch": 18.8,
"eval_loss": 0.15402325987815857,
"eval_runtime": 105.5951,
"eval_samples_per_second": 84.673,
"eval_steps_per_second": 1.326,
"step": 60000
},
{
"epoch": 18.83,
"learning_rate": 7.945275322909587e-07,
"loss": 0.0252,
"step": 60100
},
{
"epoch": 18.86,
"learning_rate": 7.73283480625425e-07,
"loss": 0.0258,
"step": 60200
},
{
"epoch": 18.89,
"learning_rate": 7.520394289598912e-07,
"loss": 0.0256,
"step": 60300
},
{
"epoch": 18.93,
"learning_rate": 7.307953772943576e-07,
"loss": 0.0253,
"step": 60400
},
{
"epoch": 18.96,
"learning_rate": 7.095513256288239e-07,
"loss": 0.0243,
"step": 60500
},
{
"epoch": 18.99,
"learning_rate": 6.883072739632904e-07,
"loss": 0.0246,
"step": 60600
},
{
"epoch": 19.02,
"learning_rate": 6.670632222977567e-07,
"loss": 0.0254,
"step": 60700
},
{
"epoch": 19.05,
"learning_rate": 6.458191706322231e-07,
"loss": 0.0248,
"step": 60800
},
{
"epoch": 19.08,
"learning_rate": 6.245751189666893e-07,
"loss": 0.0237,
"step": 60900
},
{
"epoch": 19.11,
"learning_rate": 6.033310673011557e-07,
"loss": 0.0243,
"step": 61000
},
{
"epoch": 19.14,
"learning_rate": 5.82087015635622e-07,
"loss": 0.0248,
"step": 61100
},
{
"epoch": 19.18,
"learning_rate": 5.608429639700884e-07,
"loss": 0.0251,
"step": 61200
},
{
"epoch": 19.21,
"learning_rate": 5.395989123045548e-07,
"loss": 0.0239,
"step": 61300
},
{
"epoch": 19.24,
"learning_rate": 5.183548606390212e-07,
"loss": 0.0234,
"step": 61400
},
{
"epoch": 19.27,
"learning_rate": 4.971108089734875e-07,
"loss": 0.0255,
"step": 61500
},
{
"epoch": 19.3,
"learning_rate": 4.7586675730795384e-07,
"loss": 0.024,
"step": 61600
},
{
"epoch": 19.33,
"learning_rate": 4.546227056424201e-07,
"loss": 0.0243,
"step": 61700
},
{
"epoch": 19.36,
"learning_rate": 4.333786539768865e-07,
"loss": 0.0238,
"step": 61800
},
{
"epoch": 19.4,
"learning_rate": 4.1213460231135284e-07,
"loss": 0.025,
"step": 61900
},
{
"epoch": 19.43,
"learning_rate": 3.908905506458192e-07,
"loss": 0.0238,
"step": 62000
},
{
"epoch": 19.46,
"learning_rate": 3.6964649898028556e-07,
"loss": 0.0239,
"step": 62100
},
{
"epoch": 19.49,
"learning_rate": 3.484024473147519e-07,
"loss": 0.0241,
"step": 62200
},
{
"epoch": 19.52,
"learning_rate": 3.271583956492182e-07,
"loss": 0.0241,
"step": 62300
},
{
"epoch": 19.55,
"learning_rate": 3.0591434398368456e-07,
"loss": 0.0252,
"step": 62400
},
{
"epoch": 19.58,
"learning_rate": 2.8467029231815094e-07,
"loss": 0.0242,
"step": 62500
},
{
"epoch": 19.61,
"learning_rate": 2.634262406526173e-07,
"loss": 0.0237,
"step": 62600
},
{
"epoch": 19.65,
"learning_rate": 2.421821889870836e-07,
"loss": 0.0237,
"step": 62700
},
{
"epoch": 19.68,
"learning_rate": 2.2093813732154997e-07,
"loss": 0.0242,
"step": 62800
},
{
"epoch": 19.71,
"learning_rate": 1.9969408565601633e-07,
"loss": 0.0246,
"step": 62900
},
{
"epoch": 19.74,
"learning_rate": 1.7845003399048266e-07,
"loss": 0.0244,
"step": 63000
},
{
"epoch": 19.77,
"learning_rate": 1.5720598232494902e-07,
"loss": 0.0242,
"step": 63100
},
{
"epoch": 19.8,
"learning_rate": 1.3596193065941536e-07,
"loss": 0.0243,
"step": 63200
},
{
"epoch": 19.83,
"learning_rate": 1.1471787899388172e-07,
"loss": 0.024,
"step": 63300
},
{
"epoch": 19.87,
"learning_rate": 9.347382732834808e-08,
"loss": 0.0241,
"step": 63400
},
{
"epoch": 19.9,
"learning_rate": 7.222977566281442e-08,
"loss": 0.0234,
"step": 63500
},
{
"epoch": 19.93,
"learning_rate": 5.098572399728077e-08,
"loss": 0.0238,
"step": 63600
},
{
"epoch": 19.96,
"learning_rate": 2.9741672331747115e-08,
"loss": 0.0238,
"step": 63700
},
{
"epoch": 19.99,
"learning_rate": 8.49762066621346e-09,
"loss": 0.0246,
"step": 63800
},
{
"epoch": 20.0,
"step": 63840,
"total_flos": 5.895453426057216e+20,
"train_loss": 0.07169675377377292,
"train_runtime": 490800.9612,
"train_samples_per_second": 33.299,
"train_steps_per_second": 0.13
}
],
"max_steps": 63840,
"num_train_epochs": 21,
"total_flos": 5.895453426057216e+20,
"trial_name": null,
"trial_params": null
}